In this accident, the flight crew's take-off performance calculations resulted in an error that remained undetected until the aircraft reached a point where the crew's response was too late to avert the accident. The analysis will focus on those events, conditions and/or underlying factors that were causal or contributing to the accident. Areas where there were additional and undesirable risks to the system will also be discussed.
The Halifax International Airport fire hall electrical system was configured to provide two separate sources of emergency power (generators) in case of a power failure. However, one failed to provide power because the main circuit breaker had tripped, preventing some of the associated electrical systems in the fire hall from operating as designed. The firefighters, therefore, had to respond in very dim light conditions during their initial response. Aside from creating some confusion, there was an increased risk of injury to the firefighters and the potential for a delayed response. The earlier system of self-contained battery-operated lights, which had been removed following the installation of an emergency power generator, would have provided immediate illumination of the bunk room and vehicle bay.
Grid coordinates were not used to direct the Halifax International Airport ARFF unit or any other responders at any time during the response to this accident. Because the weather conditions and visibility were good, the first responders were able to easily locate the accident site. There was, however, some confusion with other responding units as to the exact location of the accident site.
The air traffic controllers regularly assist the airport firefighters with grid run training, yet the controllers are not required to use grid coordinates when directing ARFF to an accident site. The use of grid coordinates during an emergency would prevent confusion and could reduce critical response time.
Air traffic controllers were initially unsuccessful in their attempts to get information on the number of persons and dangerous goods on board the aircraft. One hour after the accident, MK Airlines Limited provided information regarding the number of persons on board and indicated that no dangerous goods were loaded in Halifax. Since all the dangerous goods carried on MKA1602 had been loaded at the previous stop, Bradley International Airport, no one in Halifax had any information regarding the dangerous goods. It was not until 10 hours after the accident that ARFF received a listing of the dangerous goods that had been loaded at Bradley. A lack of timely information concerning dangerous goods could have jeopardized the safety of the ARFF personnel and other responding personnel. In the case of a survivable aircraft accident, knowledge of the number of occupants could be critical to successful rescue efforts.
The slope for Runway 24 was published incorrectly in the Canada Flight Supplement and the Canada Air Pilot as 0.17 per cent down. Slope information is a consideration when calculating aircraft take-off and landing performance. If published slope information is incorrect, aircraft performance data could be incorrectly calculated. It could not be determined how the 0.17 per cent value was determined. The incorrectly published runway slope was not a factor in the take-off performance of MKA1602.
NAV CANADA obtained the required approval from TC for construction of the berm at the end of Runway 24. As part of the approval process, the berm was evaluated for obstacle clearance in accordance with TC's TP 312. The berm was not considered an obstacle by definition because it did not penetrate the plane of the obstacle-free zone. Obstacles are only considered hazards if they will affect the climb profile of an aircraft that has met its certification criteria.
There are no specific standards or recommended practices regarding the construction and risk assessment of berms or similar immovable objects in the runway overrun/undershoot areas, except in the case where they constitute an obstacle or they impinge on an established RESA. TC's guidance considers a RESA only to be a recommended practice, yet ICAO considers it to be required standard. There was no RESA published for Runway 06 or Runway 24 at Halifax International Airport. The ILS localizer berms were both located beyond the minimum recommended RESA distances.
The various communications devices that emergency response personnel tried to use for communication at the crash site and between the crash site and other areas did not provide reliable communications. These communication difficulties complicated coordination during the response, and in other circumstances, could have hampered a rescue attempt or quick evacuation of an injured person.
The aircraft operating empty weight did not include crew and operational equipment carried on board the aircraft. Therefore, every time a mass and balance sheet was completed for the occurrence aircraft, the crew would not be aware that the aircraft was actually 1120 kg heavier than calculated. In some cases, this could have put the aircraft over its maximum allowable take-off or landing weight.
The ground handling agent at Halifax International Airport did not have the facilities to weigh built-up pallets that were provided by others, and because the manifest provided by the freight forwarder was believed to be accurate, an incorrect cargo manifest spreadsheet was created. The unverified cargo weight at Halifax allowed the extra weight of the wooden skids to go unaccounted for and could have also contributed to an overweight condition. As well, the failure to detect an error in the load weight could result in adverse aircraft performance and, potentially, an accident. In this case, the aircraft was still within the allowable weight and balance limits for the take-off at Halifax.
MK Airlines Limited was aware that some loadmasters had been using an unapproved electronic version of the load planning sheet for approximately 2½ years before the accident. The company did not prevent the loadmasters from using this software, even though it had not been verified that the software and database were free from errors. Therefore, there was a risk that an aircraft could have been improperly loaded without the crew's knowledge. This was an example of the company not exercising adequate oversight to correct a known adaptation of company procedures by flight crews.
MK Airlines Limited had grown significantly during its relatively short history. The company's commercial success and subsequent expansion increased demands on its infrastructure. The addition of the B747 aircraft added significantly to the Training Department's challenge of meeting the demand for qualified flight crews. At the same time, flight crew turnover was increasing as individuals found more attractive employment elsewhere. Also, the company's policy of recruiting from southern Africa limited the pool of new potential crew members. All these factors contributed to a shortage of flight crew required to meet the flying or production demand. This shortage of flight crews increased the potential for increased fatigue and stress among the personnel.
Although the OM stated that flights would not be planned beyond 24 hours, the Crewing Department at MK Airlines Limited routinely scheduled flights in excess of that limit. There was no effective program in place to monitor how frequently these planning exceedences occurred, nor was there a program to detect and monitor exceedences beyond the planned duty days. In the absence of adequate company corrective action regarding these exceedences, crews developed risk mitigation strategies that included napping in flight and while on the ground to accommodate the longer scheduled duty days. This routine non-adherence to the OM contributed to an environment where some employees and company management felt that it was acceptable to deviate from company policy and/or procedures when it was considered necessary to complete a flight or a series of flights.
There is a reasonable limit to the time a flight crew can remain on duty before acute fatigue begins to induce unacceptable human performance deficiencies. This is regardless of the crew composition and the adequacy of the rest facilities on board the aircraft. Examination of the occurrence crew's work/rest/sleep and duty history indicated that the operating crew would have been at their lowest levels of performance because of fatigue at, or shortly after, their arrival in Halifax. This state of fatigue would have made them susceptible to taking procedural shortcuts and reduced their situational awareness. This period of low performance would have been present when the take-off performance data were calculated, the before-flight SOPs were not followed, and the inadequate take-off performance was not recognized.
The company's flight and duty scheme allowed flights to be scheduled up to 24 hours with only three pilots required. This meant that there would be either only one captain or one first officer in the crew. Because most crew members were only qualified to occupy either the left or right pilot seat, two of the assigned pilots would have to be present for every take-off, departure, arrival, and landing for the entire route. This resulted in the lone captain or first officer being subjected to a disproportionate amount of flight deck duty and, therefore, more vulnerability to fatigue. For this series of flights, the first officer was the critical crew member in this respect.
The first officer had checked out of the hotel in Luxembourg at 0925 on October 13, but it is known that he was awakened earlier than 0848, perhaps as early as 0630 or 0700. It is probable that he was not in the cockpit for a few hours on the first flight, but it is unlikely that he would have slept or had a good rest because of circadian rhythm effects. As other MK Airlines Limited flight crews indicated, it was not easy to get rest on the flight to Bradley International Airport because of the time of day. The flight from Bradley to Halifax took 1 hour 9 minutes, and the first officer would have been in the cockpit during this flight. Therefore, he would likely have been the most fatigued pilot.
The aircraft was on the ground at Halifax International Airport for 1 hour 42 minutes. Twice during this time, it was noted by ground personnel that the first officer was not in the cockpit, and it was common for flight crew to nap or rest if the turnaround time was long enough. It is likely that he took a nap between the time the take-off performance data were calculated and when he was required to be back in the cockpit to prepare for the departure.
If the first officer had been sleeping while the aircraft was on the ground in Halifax, he would have been susceptible to sleep inertia for 10 to 15 minutes after waking up. As a result, he would have been less alert than usual when he first entered the cockpit, the period when the performance data would have been set from the take-off data card information. In addition, if the captain had carried out some of the first officer's pre-flight duties to allow him to sleep, this would have further removed the first officer from the cockpit environment and decreased his situational awareness.
At the time of the occurrence, MK Airlines Limited rest, duty and flight time scheme was one of the least restrictive among ICAO signatory states. The company's increase of the maximum flight duty time for a heavy crew from 20 to 24 hours also increased the potential for fatigue.
MK Airlines Limited flight crews often flew into airports with poor facilities, experienced frequent delays and equipment malfunctions, and were scheduled for lengthy duty periods, often with limited on-board rest facilities. Many of the crews, supervisors and managers were accustomed to difficulty, hardship, and overcoming challenges. The growth and success of the company also had a great deal to do with the familial environment. Unfortunately, some of the strengths that this environment brought also generated weaknesses. These weaknesses were in the form of accepting shortcuts and deviations from procedures when it was deemed appropriate. An example of this was the acceptance of non-adherence to company direction and procedures by both management and line personnel.
Acceptance of non-adherence to company direction and procedures by managers was often tacitly accepted in the belief that it did not generate an unacceptable risk. Although three previous accidents should have been significant risk indicators for the company, there was an overall acceptance that the commercial growth (production) was being managed adequately in terms of risk (protection).
Several of the operating norms that were identified were similar in nature to those in James Reason's book. For example, shortcuts (non-adherence to procedures) had become a habitual part of routine work practices. Reason states, "This gradual reduction in the systems safety margins renders it [the company] increasingly vulnerable to particular combinations of accident-causing factors."
SOPs are established for the safe and efficient operation of an aircraft and are considered to be a critical defence toward ensuring consistent and safe operational outcomes. For this crew, the SOPs were critical for ensuring that the take-off performance data were calculated correctly, and any potential errors in that data were detected before the take-off was attempted. Non-adherence to SOPs, as was shown in this accident, can have catastrophic consequences.
The company OM, which had been approved by the GCAA, contained a description of how the company was to conduct flight operations safely and within the regulations. Many areas of the OM were incomplete, out of date or inadequate. Moreover, the Operations Manager was over tasked to a point where adequate supervision and management of day-to-day flight operations was not always possible. The familial nature of the company also interfered at times with ensuring that company personnel consistently adhered to company policies and procedures.
The BLT was introduced by MK Airlines Limited without direction, assistance or approval from the GCAA. Although advisory and guidance references of the FAA and Joint Aviation Authority were used, the introduction was without adequate training and evaluation. The crew reference material was self-study and there was little direct training provided. Furthermore, the quick reference information provided in the Notice to Flight Crew of 29 March 2004 did not specifically remind pilots that, when returning from the weight and balance page, the take-off weight as listed in that page would appear in the planned take-off weight block on the performance page. This feature is believed to be a key element in how the incorrect take-off performance data were generated. It is unknown if the user(s) of the BLT in this occurrence was fully conversant with the software, in particular this feature.
In general, the safety oversight the GCAA conducted on MK Airlines Limited was limited. The GCAA's oversight effectiveness was adversely affected by the necessity to maintain a greater amount of scrutiny on another Ghana-registered airline, even though the following significant risk indicators were present at MK Airlines Limited:
The delay in obtaining Ghana parliamentary approval for new regulations and the diversity of the company's operations also affected the oversight activities. Particularly noteworthy was the undetected, ongoing and substantial exceedences related to crew flight and duty times.
In general, the regulatory oversight of MK Airlines Limited by the GCAA was not adequate to detect serious non-conformances to flight and duty times, or ongoing non-adherence to company directions and procedures.
Without a CVR, it was difficult to determine the exact reasons the flight crew used a low EPR setting and a low rotation speed; however, a comparison of the Bradley take-off performance data against the Halifax take-off performance data was very revealing. The first part of this section will discuss possibilities that were likely not a factor in the take-off performance of the aircraft.
One possibility is that the flight crew did not change the airspeed bugs from their positions after the Halifax landing. If this were true, two white bugs would have been set together at 133 knots, appearing to be a single airspeed bug position. The command speed bug would have been set at 143 knots, appearing to be the second airspeed bug position.28 This would have appeared unusual to the flight crew for a take-off because the command speed bug would usually be the third bug in order, not the second. Also, when Vr was called at the second bug, the FDR would have recorded the aircraft rotating after 143 knots. The FDR data showed the aircraft rotating at 130 knots. This scenario is unlikely.
Another possibility is that the flight crew used the Bradley take-off data card since the power settings and rotations were similar. Had the flight crew not completed their post-flight activities in Halifax and not put the Bradley take-off data card in the trip envelope, then it could have been possible to mistakenly use the Bradley take-off data card. However, the card would have had a weight of 240 000 kg written on it, with a stabilizer trim setting of 4.0 units.
It is likely that the captain would have noticed the lower weight and a different trim setting from the one used in Halifax when he compared the mass and balance sheet to the take-off data card. The trim setting that had been determined by the crew in Halifax was appropriate for the aircraft actual take-off weight at Halifax. The correct trim value for Halifax was also confirmed by examination of the FDR data, the trim indicator and the screw jack measurements. If the Bradley data card had been used in Halifax, then a flight crew member would also have had to erase or cross out the Bradley trim value and replace it with the Halifax value. This scenario is unlikely.
If the flight crew had completed a new take-off data card for the Halifax International Airport, they would have had three options to calculate the take-off data: runway analysis charts, Volume 2, or the BLT. Since the runway analysis charts had been removed from all the aircraft, this scenario is only possible if one of the flight crew members had his own personal copies; this is unlikely. The take-off speeds chart of Volume 2 does not provide for obstacle clearance. If the crew had calculated the speeds based on the obstacles by using Volume 2, it would have been time consuming because of the work involved in finding and interpreting all the graphs, tables and charts. If the crew used only the take-off speeds chart of Volume 2, they would have first written the flight planned take-off weight of 353 tonnes onto the take-off data card, then transcribed a V1 of 150 knots, a Vr of 161 knots, and a V2 of 172 knots. Also, if Volume 2 had been used, it would not explain the thrust setting of 1.33 EPR. The BLT had been in use for several months and it was reported that it was being used for take-off performance calculations. Therefore, it is unlikely that Volume 2 was used. The BLT was most likely the source used for the take-off data. Consequently, it is most likely that the performance data error came from the misuse or misunderstanding of the BLT.
There are various scenarios that would have created the erroneous data. Assuming that the user input the correct airport, runway, and atmospheric information, then the only factor that would determine the V speeds and the EPR settings is the weight of the aircraft used in the planned weight box. If the user mistakenly used the zero fuel weight (262 000 kg) or landing weight (281 000 kg), the rotation speeds would be too high compared to what was found on the FDR. Another possibility is that the user input 253 000 kg instead of 353 000 kg by mistake; again, this weight is too high for the rotation speed in Halifax recorded on the FDR. Therefore, the only weight that generates the same rotation speed and EPR settings as found in Halifax is the Bradley weight, 240 000 kg.
The user would likely transcribe the weight of 353 000 kg from the flight plan on a white take off data card. When the BLT program was launched, all the previous settings, data, and information from the last use would have been populated in all the fields. The BLT was last used at Bradley International Airport with RTG II. Therefore, once the user opened the software on the RTG II page, he would have had to change all the fields to the Halifax International Airport, runway and ATIS. If the maximum thrust rating (7Q engines) had been selected, the lowest EPR value that could have been generated by the BLT was 1.40 EPR; therefore, the RTG II page had to have been used. Anytime after that, if the user opened the weight and balance page, for whatever reason, and returned to the take-off performance page, the planned weight dialogue box would be populated with the take-off weight from the weight and balance page, that is Bradley (240 000 kg). If the user did not know about this feature or did not notice the change and selected "calculate," the V speeds and EPR setting would have been identical to those for take-off from Bradley (see Figure 6).
A more comprehensive training program for the BLT that emphasized human factors and the potential for human error as described in the guidance material, combined with a method of ensuring that individuals were competent using the software, would certainly have reduced the possibility of this type of operator error.
If the user then wrote these performance numbers on the take-off data card with the correct planned weight of 353 000 kg for Halifax, it is likely that the error would not be noticed at this point. For this error to proceed to the next level, the other crew member either did not do a cross-check or did a cross-check and made the same error.
It is highly unlikely that both crew members would make the same error. For the incorrect V speeds to be set on the ASIs, it is likely that the gross error check for the planned weight of 353 000 kg was not done, because if it had been, the error would have been detected. Once the bugs were set on both ASIs, any subsequent checks would just validate the erroneous settings. If the weight of 353 000 kg was written on the card, then it would further support the numbers being correct when the captain signed and cross-checked the mass and balance sheet, and when the flight engineer set his total weight indicator.
If the Boeing T-card or a company-amended version of the T-card had been used, the take-off weight used to generate the performance data would have been printed on the T-card and it would not have matched the aircraft weight calculated by the loadmaster and noted on the mass and balance sheet.
In this accident, the take-off was attempted using a thrust setting and take-off speeds significantly lower than those required to become safely airborne. The company's standard call for "Set MAX POWER" during every take-off would not have provided any additional opportunity to make the crew aware if the power being set was maximum or reduced. Once the take-off began, the flight crew did not recognize that the aircraft's performance was significantly less than the scheduled performance, until they reached a point where their response was insufficient to avert the accident.
Several similar accidents and incidents have shown that there have been other crews throughout the aviation industry that have not recognized inadequate take-off performance. Some of these occurrences have resulted in substantial aircraft damage and, in several cases, substantial loss of life. Notwithstanding over 30 years of effort within the industry, there still does not appear to be an acceptable industry "in-cockpit" defence that would provide crews with timely information when take-off performance is inadequate to become safely airborne.
The take-off data card was most likely completed using performance data from the BLT. The FDR data for the Halifax take-off was nearly identical to that of the Bradley take-off, indicating that the Bradley take-off weight was used to generate the performance data in Halifax. The Bradley weight in the weight and balance page was likely unknowingly transferred to the performance page due to a reversion feature of the software. The user subsequently selected "calculate," which resulted in the generation of take-off performance data containing incorrect V speeds and thrust setting for Halifax. The flight crew used the incorrect V speeds and thrust setting during the take-off attempt; however, the settings were too low, especially the thrust setting, to enable the aircraft to take off safely.
Factors that likely contributed to the incorrect take-off data being generated and then not being detected before the take-off attempt were flight crew fatigue, non-adherence to procedures, inadequate training on the BLT, and personal stresses. Once the take-off had commenced, the crew's situational awareness likely was not sufficient to allow them to detect the inadequate acceleration before it was too late to take off safely. Factors that likely contributed to this condition were flight crew fatigue and a dark take-off environment.
On 20 October 2004, the Transportation Safety Board of Canada (TSB) issued Safety Advisory A040058-1 (Verification of Cargo Weights) to Transport Canada (TC). The Safety Advisory indicated that TC might wish to examine the adequacy of cargo handling procedures, both inside and outside Canada, and, in particular, the adequacy of load weight verification and the regulatory oversight of these issues.
On 09 December 2004, TC responded to Safety Advisory A040058-1. The letter quoted several regulations applicable to commercial operations: International Civil Aviation Organization (ICAO) standards and recommended practices, Canadian Aviation Regulations, Joint Aviation Requirements (JARs), and Federal Aviation Regulations (FARs). TC stated that the intent of the regulations is to ensure that the actual weight of the cargo, including the weight of the contents, the packing material, the packaging, the pallet or unit load device, the strapping, the wrapping, and any other device or material being transported with the cargo is accounted for in the total weight of the cargo. TC further stated that the regulations clearly indicate that it is an operator's responsibility to ensure that proper weighing procedures are in place to support its operations.
It is TC's position that the existing regulations and standards adequately address the issues raised in the Safety Advisory. However, in light of the recent accident in Halifax, Nova Scotia, and to reinforce the absolute necessity for accurate load control, TC published a Commercial and Business Aviation Advisory Circular on this issue on 04 June 2005.
On 22 October 2004, the TSB issued Safety Advisory A040059-1 (Runway Slope Information - Publication Errors) to TC. The Safety Advisory raised a concern about the accuracy of published runway slope information. The slope datum for Runway 24 at Halifax International Airport published in Canadian aeronautical information publications was incorrectly depicted as 0.17 per cent down, when it should have read 0.17 per cent up. The Safety Advisory suggested that TC might wish to ensure that similar runway slope information errors do not exist for other aerodromes. A review of quality assurance measures regarding the provision and depiction of aerodrome information in Canadian flight information publications was suggested.
On 09 December 2004, TC responded to Safety Advisory A040059-1. TC indicated that a preliminary review has not uncovered further errors in published runway slope data. The error in the Halifax data is the result of a human transposition error. The error in the slope value for Runway 24 was discovered in the course of this accident investigation. TC was advised of the error, and a NOTAM (Notice to Airmen) was issued instructing holders of the Canada Air Pilot to delete the slope information for all runways at the Halifax International Airport.
Subsequently, it was discovered that, in accordance with NAV CANADA's Aeronautical Information Publication (A.I.P. Canada) specifications, only slope values greater than 0.3 per cent are published. Therefore, a second NOTAM was issued, instructing holders of the applicable documents to re-insert the slope value for Runway 33, and delete the value for Runway 24.
Additionally, as a result of this Safety Advisory, an Aerodrome Safety Urgent Bulletin was sent to TC regional offices for distribution to all airports and registered aerodromes. The Urgent Bulletin reminds all airport/aerodrome operators of their responsibility to verify the accuracy of all published data, and to report immediately, via a NOTAM, the corrections to be made to aeronautical information publications. Direction concerning the methodology for the calculation of runway slope data is being reviewed and coordinated with NAV CANADA and other interested stakeholders.
NAV CANADA subsequently advised the TSB that, between 01 January 2004 and 01 October 2004, only two requests were received to amend runway slope information contained in its aeronautical information publications. After the issue of Safety Advisory A040059-1 and up to 01 December 2005, NAV CANADA had received a total of 73 requests to amend or to add runway slope information.
In a letter dated 01 November 2004, the Ghana Civil Aviation Authority (GCAA) instructed MK Airlines Limited to cease use of the Boeing Laptop Tool (BLT) until such time as approval is given by the GCAA. Additionally, the GCAA instructed MK Airlines Limited to comply with the crew rest requirements listed in Section 8.11 of the Ghana Civil Aviation Regulations (GCARs) until the company submits a new schedule for approval; and apply Section 8.11.1.3(a)(4) of the CGARs for loadmaster rest and Section 9.4.1.16 of the CGARs for ground engineer crew rest.
On 20 October 2004, MK Airlines Limited issued a Notice to Flight Crew that stated, "Loadmasters and Station Officers are required to query weights on dead-load weight statements when two and/or several pallets are reflected at or near the same weights. Where necessary, a check weight on a suitable calibrated scale must be carried out prior to loading. DO NOT under any circumstances accept freight that has not been weighed over a calibrated and current aircraft pallet scale from a Company approved handling agent/company. Any occurrence of any nature in this regard requires an occurrence report to be completed and submitted to the FSO [Flight Safety Officer]."
On 20 October 2004, MK Airlines Limited issued a Notice to Flight Crew on the above topic that stated, "With immediate effect to avoid any confusion, the weight and index for the total number of people on board (flight deck and all passengers) must be shown in the 'correction' box in both the B747 and DC-8 load sheets. The basic weight does not include 'crew'."
Within two weeks of the accident, MK Airlines Limited issued a Notice to Flight Crew to immediately cease use of the BLT and use alternate procedures. A formal submission has been made to the United Kingdom Civil Aviation Authority (CAA) in accordance with Temporary Guidance Leaflet No. 36: Approval of Electronic Flight Bags (EFBs).
On 03 February 2005, MK Airlines Limited issued a Notice to Flight Crew on the above topic that stated that the EFBs (JeppView® computers) are not to be used until such time as their use has been formally approved in compliance with the JAR guidelines.
At the request of MK Airlines Limited, the United Kingdom government, in cooperation with the GCAA, conducted a full audit for ICAO compliance on 16 November 2004. As a result of the audit, MK Airlines Limited decided to obtain JAR compliance in accordance with the United Kingdom authorities.
The United Kingdom CAA publication entitled Avoidance of Fatigue in Air Crews (CAP 371) lists the regulations for the avoidance of fatigue in aircrew. MK Airlines Limited has asked the GCAA for approval to use this flight time limitation scheme and has amended its operations manual (OM) accordingly. This scheme was approved in May 2005 and has been fully implemented, and United Kingdom CAA inspectors are monitoring compliance. A crew notice was issued concerning the noting of duty times on voyage reports to enable better monitoring of required rest times.
The MK Airlines Limited rostering staff has been briefed on the CAP 371 limitations and will monitor crew scheduling with in-house developed software to prevent exceedences. Crews were briefed on the new flight time limitations and their responsibilities for compliance. Flight documents are subject to close inspection to ensure that captain's discretion reports are completed when required.
A crew notice was issued concerning counselling to reduce fatigue and stress in light of the accident and the continued political and security situation in southern Africa. A new pay scheme introduced in December 2004 improved the financial security of crew members and has been well received.
The audit program of ground service contractors has been enhanced by the qualification of MK Airlines Limited loadmasters to the British Standards Institution (BSI) Lead Assessor standard. More detailed audit procedures have been developed under control of the newly appointed Director, Safety and Quality. The new Safety and Quality Department will coordinate audit activities across the technical, operational, security, and traffic disciplines.
A safety management system was established throughout the company, and a new company safety policy was drafted. A program of flight data monitoring is being implemented as part of the flight operations quality system. This forms part of the safety management system, which integrates safety, quality, and security management of the company. Key staff attended a flight operations quality assurance course from 04 April to 07 April 2005, at Cranfield University, United Kingdom.
A safety culture questionnaire was drafted and included in the latest company safety magazine to acquire employee feedback.
There have been extensive revisions to the training manual under United Kingdom CAA guidance to achieve JAR compliance. A new Assistant Training Manager was appointed, who has an extensive background in training management. Training is now compliant with JAR-FCL (flight crew licensing)/JAR-OPS (operations) requirements. Non-JAR-licensed aircrew have commenced study courses for JAR licences.
Numerous companies and consultants were contracted or employed to oversee training standards:
The company's OM (Part A, Section 8 - Draft for approval by United Kingdom CAA) has been updated with various flight briefings to improve the level of situational awareness. Procedures were developed to ensure continued alignment of company training manuals with current national and international regulations and manufacturer's service bulletins. The training record system for each fleet was reviewed and aligned with the requirements of the training manual. A training expiry database was developed and will be reviewed to ensure that Part D, JAR-FCL requirements are met. Airworthiness directives and manufacturers service bulletins procedures are in Section 2.11 of the maintenance control manual and are being rewritten in the Maintenance Organization Exposition to comply with Section 145 of the JARs. An Information Management Department was established.
A system has been developed to improve the crew qualification system for categories B and C aerodromes. Crew participation in information gathering and updating of aerodrome information has been implemented, incorporating a risk assessment and controlled flight into terrain analysis. Part C of the OM now includes route and aerodrome briefs. In addition to the air operating certificate (AOC) application, MK Airlines Limited will apply for Type Rating Training Organization (TRTO) in accordance with the JARs. MK Airlines Limited is in the process of publishing a TRTO manual for approval, which will contain all the requirements for instructors, instructor training, standards, and facilities.
MK Airlines Limited has initiated a process to ensure that all company instructors and examiners are qualified to JAR-FCL, Subpart I. For example:
The first Boeing 747 training course, in accordance with the new JARs, commenced in May 2005. Work is under way at the company operations centre at Landhurst to add a Training Department wing to the existing structure. The new wing will house the Boeing 747 simulator, the Training Department personnel offices, and the Safety and Quality Department.
Proficiency check rides for each flight crew member have been updated to reflect the requirements of sections 1.240, 1.295 and 4.240 of JAR-CFL, Appendix 2. MK Airlines Limited has initiated a biennial technical ground training program for aircraft systems. Syllabus and course material comply with Subpart N of JAR-OPS. MK Airlines Limited has completed a program of initial and recurrent training for loadmasters to include crew resource management training. The chief loadmaster attended CRMI training from 10 January to 14 January 2005. Eight 3-day courses were conducted for all loadmasters between 16 January and 12 February 2005.
Enhanced ground proximity warning systems (EGPWS) have been added to the ground and simulator training for the B747 fleet. A two-hour EGPWS lecture was programmed for January to June 2005 during recurrent training. An interactive CD ROM was issued to each crew member for home study.
A biennial training requirement for aircraft performance and de-icing was added to the recurrent ground training program. De-icing holdover times in the OM, Part A, Section 8.2.4.8.3, were amended on 21 December 2004 to bring them in accordance with the Association of European Airlines tables (19th edition, September 2004). A Crew Notice was issued on 21 December 2004 to ensure crew awareness and to provide interpretation information.
MK Airlines Limited has applied for a United Kingdom AOC.
The company organization has been revised to comply with or surpass the JARs, with the addition of several new, full-time appointments of highly qualified staff. The AOC project plan is managed by the Director, Safety and Quality and was completed in late 2005.
In support of the AOC application, the following activity has been initiated and/or completed:
A revised B747 JAR-compliant minimum equipment list, based on the United Kingdom CAA master minimum equipment list, was submitted for approval. Changes were required and the amended version has been resubmitted as part of the AOC application.
The newly appointed Ground Operations Manager has implemented tighter oversight of contracted ground handling companies.
A full check of aircraft equipment by the United Kingdom CAA was undertaken and confirmed that all aircraft were fully ICAO-compliant. Fuelling and loading procedures were reviewed to ensure standardization across both B747 and DC-8 fleets.
In February 2005, the company commenced the application process for European Aviation Safety Agency (EASA) Part M (maintenance management) and EASA Part 145 approval. In March 2005, MK Airlines Limited started the recruitment of EASA Part 66 licensed staff and introduced training for existing staff to convert licences. In the first six months of 2005, the company undertook or completed several safety actions regarding airworthiness:
The Transportation Safety Board of Canada (TSB) investigated this occurrence for the purpose of advancing transportation safety. It is not the function of the Board to assign fault or determine civil or criminal liability.
Aviation Investigation Report
Reduced Power at Take-off and
Collision with Terrain
MK Airlines Limited
Boeing 747-244SF 9G-MKJ
Halifax International Airport, Nova Scotia
14 October 2004
Report Number A04H0004
On 14 October 2004, an MK Airlines Limited Boeing 747-244SF (registration 9G-MKJ, serial number 22170) was being operated as a non-scheduled international cargo flight from Halifax, Nova Scotia, to Zaragoza, Spain. At about 0654 coordinated universal time, 0354 Atlantic daylight time, MK Airlines Limited Flight 1602 attempted to take off from Runway 24 at the Halifax International Airport. The aircraft overshot the end of the runway for a distance of 825 feet, became airborne for 325 feet, then struck an earthen berm. The aircraft's tail section broke away from the fuselage, and the aircraft remained in the air for another 1200 feet before it struck terrain and burst into flames. The aircraft was destroyed by impact forces and a severe post-crash fire. All seven crew members suffered fatal injuries.
Ce rapport est également disponible en français.
©Minister of Public Works and Government Services 2006
Cat. No. TU3-5/04-2E
ISBN 0-662-43546-X
Time (UTC) | Sequence of Events Summary | Indicated Airspeed (knots) | Pitch (degrees) | Distance from Runway 24 threshold (feet) |
---|---|---|---|---|
0647:06 | Flight data recorder (FDR) powers on. All engines are running, parking brake is on, flaps are retracted, horizontal stabilizer is at 3.3 trim units, heading is 337ºM, static air temperature is 7.7ºC | 50 | -0.9 | |
0648:36 | Thrust levers are advanced | 50 | -0.9 | |
0648:40 | Aircraft begins to taxi | 50 | -0.9 | |
0648:58 | Flap handle position set to 20º. Horizontal stabilizer position briefly moves from 3.28 trim units to 4.9, then briefly decreases to 4.6, followed by further increase to 6.1 trim units (and remains at 6.1 trim units to end of flight) | 50 | -0.4 | |
0649:05 | Leading edge flaps extended | 50 | -0.4 | |
0649:48 | Aileron control check performed | 50 | -0.4 | |
0650:03 | Elevator control check performed | 50 | -0.4 | |
0650:14 | Rudder control check performed | 50 | -0.4 | |
0651:13 | Starts onto runway | 50 | 0 | |
0651:51 | Backtracks Runway 24 | 50 | -0.9 | |
0652:49 | Starts right turn at threshold, to line up for take-off Runway 24 | 50 | -0.9 | |
0653:18 | Stops right turn on heading 240º, aligned with runway centreline | 50 | -0.4 | 194 |
0653:19 | Thrust levers are advanced for rolling take-off | ? | -0.4 | 205 |
0653:31 | Split occurs in throttle lever angles | ? | -0.4 | 474 |
0653:35 | Recorded airspeed begins to record data above 50 knots indicated airspeed (KIAS) | 51 | -0.5 | 725 |
0653:36 | Engines stabilize at take-off power (engine pressure ratios at 1.32 to 1.34; engines 2 & 3 thrust lever angles (TLAs) at 47 per cent and 43 per cent, while engines 1 & 4 TLAs at 30 per cent) | 53 | -0.5 | 801 |
0654:08 | Control column moves aft to initiate rotation | 130 | -0.5 | 5483 |
0654:10 | Nose-up rotation commences; elevator data mirrors control column movement | 135 | -0.4 | 5907 |
0654:13 | Control wheel moves clockwise to 6.1º (6º to 7º clockwise input for next 6 seconds); control column moves aft to 8.4º | 140 | 2.3 | 6571 |
0654:15 | Rudder (lower) deflects right to 2.5º | 143 | 6.7 | 7026 |
0654:16 | Control column moves to 8.3º aft; pitch briefly stabilizes at 9º; pitch rate approximately 2.2 degrees per second | 143 | 8.5 | 7257 |
0654:17 | Control column moves further aft to 10º (data loss due to tape splice) | 145 | ? | 7490 |
0654:18 | Control column moves to 9.1º aft; pitch reaches 10.6º (next 4 samples indicate pitch stabilizing in 11º range, consistent with lower aft fuselage/ground contact) | ? | ? | 7726 |
0654:19 | Control column at 9.0º aft; pitch reaches 11.1º | ? | 10.7 | 7965 |
Start of initial runway scrape mark number 1 | 7977 | |||
End of scrape mark number 1 | 8103 | |||
0654:20 | Control wheel moves clockwise to 14.2º; control column moves aft to 12.0º; pitch reaches 11.5º; thrust levers are advanced | 149 | 11.2 | 8207 |
Start of second runway scrape mark number 2 | 8389 | |||
0654:21 | Control column moves aft to 12.6º; pitch at 11.5º; engine pressure ratios of engines 2 & 3 reach 1.6 (maximum commanded thrust); engine pressure ratios of engines 1 & 4 are no longer recorded | 149 | 11.5 | 8449 |
0654:22 | Control column moves further aft to 13.5º; rudder (lower) deflects right to peak 8º; pitch reaches 11.9º (data loss likely due to lower aft fuselage contact with runway) | 152 | 11.6 | 8692 |
End of useable runway | 8800 | |||
0654:23 | FDR data loss | ? | ? | 8939 |
0654:24 | Control column at 13.4º aft; pitch reaches 14.5º; first tilt switch sample indicating “Air” | ? | 14.5 | 9188 |
0654:25 | Data loss likely due to lower aft fuselage impact with localizer berm; localizer berm at 9955 feet; therefore, position discrepancy equates to an error of approximately 5 per cent | 155 | ? | 9438 |
End of scrape mark number 2 in grass | 9622 | |||
0654:26 | Data loss due to localizer berm strike continues for another second | ? | ? | 9691 |
0654:27 | Data loss | ? | ? | 9947 |
Localizer berm position | 9955 | |||
0654:28 | Maximum recorded radio altitude is 36 feet; pitch decreases to -5.4º | ? | ? | 10 206 |
0654:29 | Final recorded sample of radio altitude at 0 feet; pitch further decreases to -20.9º; horizontal stabilizer position records spurious value of -72.8 trim units following localizer berm impact; FDR stops recording | R | -8.8 | 10 468 |
AC | Advisory Circular |
---|---|
A/C | aircraft |
AFM | aircraft flight manual |
agl | above ground level |
AOC | air operating certificate |
ARFF | Aircraft Rescue and Fire Fighting |
ASI | airspeed indicator |
asl | above sea level |
ATC | air traffic control |
ATIS | automatic terminal information service |
ATPL | airline transport pilot licence |
B747 | Boeing 747-200 |
BLT | Boeing Laptop Tool |
BSI | British Standards Institution |
CAA | Civil Aviation Authority |
CAP | Civil Air Publication |
CAP 371 | United Kingdom Civil Aviation Authority publication entitled Avoidance of Fatigue in Air Crews |
CRMI | Crew Resource Management Instructor |
CVR | cockpit voice recorder |
EASA | European Aviation Safety Agency |
EFB | electronic flight bag |
EGPWS | enhanced ground proximity warning system |
EPR | engine pressure ratio |
ERS | emergency response services |
FAA | Federal Aviation Administration (United States) |
FARs | Federal Aviation Regulations |
FCL | flight crew licensing |
FDR | flight data recorder |
FL | flight level |
FSO | Flight Safety Officer |
GCAA | Ghana Civil Aviation Authority |
GCARs | Ghana Civil Aviation Regulations |
GMF | Garuda Maintenance Facilities |
HIAA | Halifax International Airport Authority |
HRM | Halifax Regional Municipality |
IASA | International Aviation Safety Assessments |
ICAO | International Civil Aviation Organization |
ILS | instrument landing system |
in Hg | inches of mercury |
JARs | Joint Aviation Requirements |
KCAS | knots calibrated airspeed |
kg | kilograms |
KIAS | knots indicated airspeed |
LUX | Luxembourg |
m | metres |
MAC | mean aerodynamic chord |
MKA1601 | MK Airlines Limited Flight 1601 |
MKA1602 | MK Airlines Limited Flight 1602 |
N | north |
N/A | not applicable |
NOTAM | Notice to Airmen |
NTSB | National Transportation Safety Board (United States) |
OM | operations manual |
PDUs | power drive units |
PF | pilot flying |
QRH | quick reference handbook |
RCMP | Royal Canadian Mounted Police |
RESA | runway end safety area |
RTG II | Rating II |
SARPs | standards and recommended practices |
SITA | Société Internationale de Télécommunications Aéronautiques |
sm | statute miles |
SOPs | standard operating procedures |
STAS | Standard Take-off Analysis Software |
TAT/EPRL | true air temperature/engine pressure ratio limit |
TC | Transport Canada |
TLAs | thrust lever angles |
TODAs | take-off distances available |
TP | Transport Publication |
TP 312 | Transport Canada Publication entitled Aerodrome Standards and Recommended Practices |
TRE | Type Rating Examiner |
TRI | Type Rating Instructor |
TRTO | Type Rating Training Organization |
TSB | Transportation Safety Board of Canada |
UTC | coordinated universal time |
VAR | visual-aural range |
V1 | take-off decision speed |
V2 | take-off safety speed |
Vmcg | minimum control speed, ground |
Vmu | expected minimum unstick speed |
Vr | rotation speed |
Vref | landing reference speed |
W | west |
Z | Zulu time (equivalent to UTC) |
' | minutes |
'' | seconds |
º | degrees |
ºC | degrees Celsius |
ºM | degrees magnetic |
ºT | degrees true |
1. See Glossary at Appendix E for all abbreviations and acronyms.
2. The term "augmented" flight crew is more commonly used in international organizations and regulations.
4. Runway analysis charts are paper-based references carried on board the aircraft and are used to calculate take-off performance for a specific runway at a particular airport. They allow the pilot to obtain take-off data and take into account atmospheric conditions, the runway condition, and obstacles in the take-off flight path.
5. Volume 2 contains graphs, tables, and charts used to calculate aircraft performance data. It also contains a one-page, quick reference table for the calculation of take-off speeds. This table does not provide information relative to obstacle clearance.
6. The aircraft was originally constructed as a B747BC (passenger/cargo combination) freighter and was subsequently converted to a B747SF (full freighter) in 1995.
7. The BLT Administrator's Guide, page 40, defines operating empty weight as the weight typically found on the aircraft during normal operations, such as flight crew plus the weight derived from an aircraft weighing.
8. When water injection is used, the cooling effects of the water on the engine enable longer engine component life and thereby permit the operator to increase the thrust.
9. The thrust produced by the JT9D engines is indicated in the cockpit as EPR. The EPR is a ratio of the pressure of the air entering the engine air inlet to the discharge pressure at the engine jet nozzle.
10. MK Airlines Limited procedures required that "max thrust" be called for all take-offs, even if a de-rated or reduced thrust setting was to be used.
11. Vmu is the calibrated airspeed at and above which the aircraft can safely lift off the ground and continue the take-off.
12. Rotation speed is the speed at which the pilot starts to pull back on the yoke to rotate the aircraft in pitch.
13. NAV CANADA is responsible for providing aeronautical information services for Canada, including runway slope information.
14. SITA - Société Internationale de Télécommunications Aéronautiques
15. Zulu is equivalent to UTC.
16. These are FDR indicated data.
17. The air/ground logic of the FDR recorded "tilt switch discrete" is determined by the main landing gear tilt indication on at least one wing or body gear on each side of the aircraft. The tilt indication is satisfied when the wing gear and body gear tilt 53 and 8 , respectively, with respect to the oleos.
18. FDR data indicated 6.1 units and the mass and balance sheet indicated 5.8 units. Differences in values are considered to be within reasonable tolerances to be consistent with a cockpit setting of 5.8 units.
19. A beta version is a version of the software to be used for operational testing before official release.
20. For clarity and consistency, the report uses the term "runway analysis charts" rather than "airport analysis charts."
21. Vref is the minimum speed at the 50-foot height in a normal landing. This speed is equal to 1.3 times the stall speed in the full-flap landing configuration.
22. TAT/EPRL computes the EPR limit for engine rating and mode selected.
23. The date April 2002 was in the document. It was, inadvertently, not updated.
24. The MK Airlines Limited SOPs were to call "MAX thrust" for all take-offs, regardless of the thrust setting being used.
25. The term "acute fatigue" is more commonly used by the TSB.
26. In accordance with the MK Airlines Limited OM, the duty period commences one hour before scheduled departure and ends 15 minutes after arrival at the ramp.
27. Professor James Reason is one of the world's leading academics in the field of understanding human error.
28. In rare cases, when the aircraft is very light, the V1 and Vr bugs can be touching because the speeds are so close. However, in this case, since the crew changed the EPR bugs, it would have been natural to change the ASI bugs.
This report was prepared in accordance with Transportation Safety Board (TSB) standards for investigation reports. In keeping with these standards, the report is organized into the following main parts:
The report can be viewed in the following formats:
To obtain additional copies of the report, please contact
TSB Communications Division
Place du Centre
200 Promenade du Portage
4th Floor
Gatineau, Quebec K1A 1K8
Canada
Telephone: (819) 994-3741
Fax: (819) 997-2239
E-mail: communications@bst-tsb.gc.ca
The series of flights for this crew originated at Luxembourg-Findel Airport, Luxembourg, on 13 October 2004, as MK Airlines Limited Flight 1601 (MKA1601),1 destined to Bradley International Airport, Windsor Locks, Connecticut, United States. The aircraft operated as MK Airlines Limited Flight 1602 (MKA1602) from Bradley International Airport to Halifax International Airport, Nova Scotia, and was to continue as MKA1602 to Zaragoza, Spain, and return to Luxembourg.
The flights were operating with a heavy crew,2 comprised of two captains, one first officer, and two flight engineers. A loadmaster and a ground engineer were also on board. The crew members for MKA1601/MKA1602 arrived at Luxembourg-Findel Airport at different times and dates. The captain and first officer of MKA1601, and flight engineer of MKA1602 operated a flight from Nairobi, Kenya, to Luxembourg-Findel Airport on October 12. The captain of MKA1602 and flight engineer of MKA1601 arrived in Luxembourg from Johannesburg, South Africa, on October 12 as operating crew of their first flight after a two-week period off duty. On October 13, the ground engineer and loadmaster arrived at Luxembourg as crew on the occurrence aircraft.
The planned departure time for MKA1601 was 1000 coordinated universal time (UTC).3 At 0848, just before the crew's departure from the hotel in Luxembourg, the MKA1601 captain received a phone call from the MK Airlines Limited station liaison officer in Luxembourg, advising of a delay to the planned departure time due to the late arrival of the aircraft and late preparation of the cargo.
The captain, first officer, and flight engineer of MKA1601 checked out of the hotel at 0925. At 0941, the captain was advised that the aircraft loading was under way, and the captain, first officer, and flight engineer proceeded to the airport. The captain and flight engineer of MKA1602 checked out of the hotel at 1052 and proceeded to the airport.
When the MKA1601 captain arrived at the airport, he received the flight documentation from the Luxembourg station liaison officer. The flight documentation was prepared by the MK Airlines Limited operations centre in Landhurst, East Sussex, United Kingdom. It included the flight brief, the trip schedule, flight routing, weather, flight plan, planned fuel requirements, and planned payload. After the captain reviewed the flight documentation, he requested that 4000 kilograms (kg) of cargo be offloaded to carry additional fuel. The crew made the necessary adjustments to their flight documentation.
Another delay developed when the loadmaster noted that some of the pallets were contaminated with soil and would not be accepted by the authorities at Bradley International Airport. A vacuum cleaner was obtained and the MK Airlines Limited station liaison officer and loadmaster began to clean the pallets. So as not to delay the flight unnecessarily, the loadmaster took the vacuum cleaner on board to finish cleaning the pallets en route. The flight departed at 1556.
The first officer was identified as the pilot communicating with air traffic control (ATC) for the flight from Luxembourg to Bradley, except for a three-hour period during which time the voice communicating with ATC was that of another crew member. MKA1601 landed at Bradley International Airport at 2322.
At Bradley International Airport, all the cargo from Luxembourg-Findel Airport was offloaded. However, the cargo loading at Bradley was prolonged due to unserviceabilities with the aircraft's cargo loading system. With a captain and flight engineer crew change, MKA1602 departed Bradley International Airport for Halifax International Airport at 0403 on October 14, carrying another delay. The MKA1602 captain was the pilot communicating with ATC; the first officer was the pilot flying (PF).
MKA1602 landed on Runway 24 at Halifax International Airport at 0512 and taxied to the ramp. After shutdown, loading of the aircraft was started. During the loading, two MK Airlines Limited crew members were observed sleeping in the upper deck passenger seats. After the fuelling was complete, the ground engineer checked the aircraft fuelling panel and signed the fuel ticket. The aircraft had been uploaded with 72 062 kg of fuel, for a total fuel load of 89 400 kg. The ground engineer then went to the main cargo deck to assist with the loading.
Once the loading was complete, the ramp supervisor for the ground handling agent went to the upper deck to retrieve the MKA1602 cargo and flight documentation. While the loadmaster was completing the documentation, the ramp supervisor visited the cockpit and noted that the first officer was not in his seat. Approximately 10 minutes later, the ramp supervisor, with the documentation, left the aircraft. At 0647, the crew began taxiing the aircraft to position on Runway 24, and at 0653, the aircraft began its take-off roll. See Section 1.11.4 of this report for a detailed sequence of events for the take-off.
During rotation, the aircraft's lower aft fuselage briefly contacted the runway. A few seconds later, the aircraft's lower aft fuselage contacted the runway again but with more force. The aircraft remained in contact with the runway and the ground to a point 825 feet beyond the end of the runway, where it became airborne and flew a distance of 325 feet. The lower aft fuselage then struck an earthen berm supporting an instrument landing system (ILS) localizer antenna. The aircraft's tail separated on impact, and the rest of the aircraft continued in the air for another 1200 feet before it struck terrain and burst into flames. The final impact was at latitude 44º52'51" N and longitude 063º30'31" W, approximately 2500 feet past the departure end of Runway 24, at an elevation of 403 feet above sea level (asl). The aircraft was destroyed by impact forces and post-crash fire. All persons on board (seven crew members) were fatally injured.
Crew | Passengers | Others | Total | |
---|---|---|---|---|
Fatal | 7 | - | - | 7 |
Serious | - | - | - | - |
Minor/None | - | - | - | - |
Total | 7 | - | - | 7 |
The first damage to the aircraft occurred when, on rotation, the aircraft's lower aft fuselage struck the runway twice and remained on the ground to a point 825 feet beyond the end of Runway 24. Severe damage occurred when the aircraft's lower aft fuselage struck the berm and the vertical stabilizer and both horizontal stabilizers separated from the fuselage. The final impact was in a wooded area where impact forces and an extensive post-crash fire destroyed the remaining aircraft structure forward of the aft pressure bulkhead (see Photo 1).
Grass was uprooted in the area beyond the end of the runway where the aft section of the aircraft fuselage had dragged on the ground; as well, a number of approach lights for Runway 06 were destroyed. The ILS localizer antenna structure sustained significant damage when the aircraft struck the berm. Telephone and power lines adjacent to the main crash site were severed just before final impact. The surrounding wooded area was heavily damaged by the post-crash fire. Unburned fuel contaminated the soil in the immediate area of the crash site, requiring an extensive environmental clean-up.
The operating flight crew of MKA1602 consisted of one captain, one first officer, and one flight engineer. The captain and flight engineer of MKA1601, a ground engineer, and a loadmaster were also on board.
Operating Flight Crew | |||
---|---|---|---|
Captain | First Officer | Flight Engineer | |
Licence | Airline Transport | Airline Transport | Flight Engineer |
Medical Expiry Date | 01 July 2005 | 17 August 2005 | 13 August 2005 |
Total Flying Hours | 23 200 | 8537 | 2000 |
Hours Last 90 days | 254 | 245 | 186 |
Hours on Type Last 90 Days | 254 | 245 | 186 |
Hours off Duty Prior to Work | 29 | 17 | 17 |
Non-Operating Crew | ||||
---|---|---|---|---|
Captain | Flight Engineer | Ground Engineer | Loadmaster | |
Licence | Airline Transport | Flight Engineer | Maintenance | Not required |
Medical Expiry Date | 15 July 2005 | 27 January 2005 | Not required | Not required |
Total Flying Hours | 6000 | 1991 | Unknown | Unknown |
Hours Last 90 Days | 171 | 202 | Unknown | 421 |
Hours on Type Last 90 Days | 171 | 202 | Unknown | Unknown |
The pilot-in-command (operating captain) of MKA1602 held a Ghanaian airline transport pilot licence (ATPL) with a valid instrument rating. He was qualified and certified in accordance with the Ghana Civil Aviation Regulations (GCARs). His licence was annotated with the remark "holder to wear spectacles which correct for near vision and shall have available a second pair whilst exercising the privileges of the license." Based on a review of the captain's medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance during the flight.
The captain had been with the company since its inception and started flying the McDonnell Douglas DC-8 with MK Airlines Limited in 1990. He was in one of the first groups of company pilots to transition to the Boeing 747-200 (B747). The captain successfully completed his United States Federal Aviation Administration (FAA) type rating training on the B747 in 1999 at the Pan Am Training Center in Miami, Florida. The captain's total flying time on the B747 was approximately 4000 hours.
In 2000, the company changed its B747 standard operating procedures (SOPs) and required all B747 flight engineers and pilots to undergo additional training. During this additional training, the captain had some difficulties adjusting to the new SOPs and his training was suspended. After two weeks of review and study, the captain returned to training and completed the course without further difficulty. Records indicate that there were instances where supervisory pilots had to counsel the captain regarding non-adherence to SOPs; however, in the period before the accident, he had demonstrated a marked improvement.
The captain trusted other crew members to perform their duties with minimal supervision. He was not comfortable using personal computers and software, such as the Boeing Laptop Tool (BLT) (see Section 1.18.1 of this report). He was more comfortable using manual methods to complete performance calculations, such as using runway analysis charts4 or Volume 25 of the aircraft flight manual (AFM). Generally, those who flew with him reported that he was competent flying the aircraft. He was respected and exercised adequate command authority in the aircraft, although he preferred to work in a casual manner.
The first officer held a Ghanaian ATPL with a valid instrument rating. He was qualified and certified in accordance with the GCARs. His last medical was conducted on 17 August 2004 with no annotations on the licence, although the medical records indicated that spectacles were worn for the eye test. The previous medical assessments were annotated with the remark "holder to wear spectacles which correct for distant vision and shall have available a second pair whilst exercising the privileges of the license." Based on a review of the first officer's medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance during the flight.
The first officer was reported to be a competent pilot and comfortable using personal computers. As the only first officer for the series of flights, he would have had to be an active crew member on duty on the flight deck for all take-offs, departures, arrivals, and landings for the series of flights.
The flight engineer's licence was valid until 12 August 2005 and was endorsed for B747 aircraft. He was qualified and certified in accordance with the GCARs. His last medical was completed on 13 August 2004 and, based on a review of his medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance during the flight.
The loadmaster was trained and qualified in accordance with company standards. Although a flight medical was not required in a licensing capacity, the loadmaster completed a company medical on 16 September 2000. He was found fit for employment and, based on a review of his medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance. Records indicate that the loadmaster had flown 421 hours on MK Airlines Limited aircraft during the previous 90 days.
The non-operating captain held a Ghanaian ATPL with a valid instrument rating. He was qualified and certified in accordance with the GCARs. His licence was annotated with a requirement for corrective lenses. His last medical was conducted on 15 July 2004 and he was found fit for duty. Based on a review of his medical records, there was no indication of any pre existing medical condition or physiological factors that would have adversely affected his performance. The non-operating captain was the pilot-in-command during the flight from Luxembourg-Findel Airport to Bradley International Airport.
The non-operating flight engineer's licence was valid until 26 January 2005 and was endorsed for B747 aircraft. He was qualified and certified in accordance with the GCARs. His last medical was completed on 27 January 2004 and, based on a review of his medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance.
The ground engineer held a Ghanaian maintenance licence endorsed for B747 aircraft. The ground engineer was not subject to a medical for licensing purposes. During his last company medical, he was found fit and, based on a review of his medical records, there was no indication of any pre-existing medical condition or physiological factors that would have adversely affected his performance.
Manufacturer | The Boeing Company |
---|---|
Type and Model | B747-244SF6 |
Year of Manufacture | 1980 |
Serial Number | 22170 |
Certificate of Airworthiness | Issued 03 May 2004; valid until 02 May 2005 |
Total Airframe Time/Cycles | 80 619 hours/16 368 cycles |
Engine Type (number of) | Pratt & Whitney JT9D-7Q (4) |
Maximum Allowable Take-off Weight | 377 842 kg |
Recommended Fuel Type(s) | Jet A, Jet A-1 |
Fuel Type Used | Jet A-1 |
On 08 October 2004, the number 2 and number 3 engines were replaced. The throttles for the number 2 and number 3 engines were significantly staggered from the number 1 and number 4 engines at reduced thrust power settings. This defect was written in the aircraft's logbook.
The most recent calculations for the occurrence aircraft's weight and centre of gravity were conducted after a C-check in Jakarta, Indonesia, on 18 September 2004. A review of the 9G-MKJ Aircraft Weight and C.G. Determination document produced by Garuda Maintenance Facilities (GMF) AeroAsia, of the Garuda Indonesia Group, indicated an aircraft basic empty weight of 157 977.5 kg and an empty centre of gravity of 32.50 per cent mean aerodynamic chord (MAC).
A review of a duplicate copy of the BLT software for the occurrence aircraft, weight and balance summary page, indicated that the operating empty weight7 was 157 977 kg; this was actually the basic empty weight of the aircraft. The BLT also indicated that the empty centre of gravity arm was 32.3 per cent MAC; this varied slightly from the 9G-MKJ Aircraft Weight and C.G. Determination document produced by GMF AeroAsia, which indicated 32.50 per cent MAC.
The occurrence aircraft carried a spares kit (also known as a fly-away kit) on board at the time of the accident flight. The kit contained spare aircraft parts and tools; MK Airlines Limited estimated the weight of the kit to be 800 kg. The aircraft also carried approximately 50 kg of catering for the crews. MK Airlines Limited used standard weights for the weight of the flight crew in the cockpit, totalling 270 kg. None of these three weights, which totalled 1120 kg, had been included in the operating empty weight in the BLT, or the mass and balance sheet that was used to calculate the aircraft weight for take-off.
The occurrence aircraft had a number of cargo floor power drive units (PDUs) removed from the aircraft and blanked off because they were unserviceable. As part of the cargo load, a large roll of steel was placed on a 20-foot-long pallet for a total weight of 13 206 kg. When the steel was being loaded onto the aircraft, it could only be moved by the cargo loading system as far as the functioning PDUs would permit. Normally, pallets can be manhandled into position if the PDUs are unserviceable, but, because of the weight of this pallet, it could only be loaded into positions LR and MR (see Figure 1).
The weight limits for positions LR and MR are 4264 kg respectively, for a total weight limit of 8528 kg. The weight of the steel and the pallet exceeded the limits by 4678 kg. The MK Airlines Limited operations manual (OM), Part A, Chapter 8, stated in part that "the loadmaster/captain must comply with additional structural limits as specified in the loading manual with regard to the maximum mass per cargo compartment." If all the PDUs had been serviceable, the steel load could have been properly placed in positions HR and JR, where the limit was 13 608 kg.
In addition to the cargo loaded at Bradley International Airport, the aircraft was loaded with 5921 US gallons of Jet A fuel. The take-off mass for Bradley was 239 783 kg, with an MAC of 25 per cent and a stabilizer trim setting of 4.0 units. The lateral imbalance was 18 248 kg, which was within allowable limits. The aircraft was within the centre of gravity limits of 13 to 35 per cent MAC for that weight.
The cargo uploaded in Halifax was comprised of 18 cargo pallets. On 13 October 2004, a local freight forwarder delivered these pallets to the MK Airlines Limited cargo handling agent at Halifax International Airport. Each pallet contained hundreds of individual STYROFOAMTM packages of fresh seafood, supported on wooden skids and secured by a cargo net. The cargo handling agency created a cargo manifest spreadsheet for the flight by taking the gross weight of each pallet, which had been supplied on the cargo manifests by the local freight forwarder. The agency then added 130 kg tare weight for the weight of the pallet and netting, for the total gross weight per pallet. There were 86 wooden skids supporting the fresh seafood on the cargo pallets. The weight of the wooden skids was not accounted for in the cargo pallet gross weight provided by the local freight forwarder, nor in the cargo manifest spreadsheet. Generally, wooden skids weigh between 20 and 25 kg; therefore, approximately 1900 to 2000 kg of extra weight was not accounted for in the cargo manifest. The local freight forwarder did not weigh the built-up pallets nor did the ground handling agent at Halifax International Airport have the facilities to weigh built-up cargo pallets that were provided by others.
In addition to the cargo loaded at Halifax, the aircraft was loaded with 88 637 litres of Jet A-1 fuel. The mass and balance sheet left behind at Halifax by the crew of MKA1602 indicated a total ramp fuel of 90 000 kg, a take-off mass of 350 698 kg with an MAC of 23 per cent, and a stabilizer trim setting of 5.8 units. The lateral imbalance was 88 kg. The aircraft was within the centre of gravity limits for that weight. The company pre-planned flight documentation indicated a minimum of 86 690 kg of fuel and a planned cargo load of 109 920 kg for a planned take-off mass of 353 310 kg.
When the weight of the wooden skids (2000 kg) and the combined weight of the fly-away kit, catering, and the flight crew (1120 kg) were added to the 350 698 kg weight calculated by the crew, the actual aircraft weight would have been approximately 353 800 kg.
The B747-200 was originally certified in 1971 with JT9D-7 engines, which had a maximum thrust of 46 300 pounds (dry) and 47 900 pounds when using water injection (wet)8 on take-off. In 1979, the JT9D-7Q engine was certified for use on the B747-200. It had a maximum thrust of 53 000 pounds; the occurrence aircraft was equipped with JT9D-7Q engines.
The maximum thrust available to an engine is dependent on the air density (pressure altitude and temperature of the air) in which the engine is operating. The maximum thrust that can be used for take-off is provided in the approved AFM, and before every take-off, the flight crew must calculate the power setting of the engine to achieve the maximum thrust. To extend engine life, it is common practice to use de-rated or reduced thrust, or a combination of both, for take offs when maximum thrust is not required, such as when taking off from long runways or with light loads.
De-rated thrust is a take-off thrust level less than the maximum take-off thrust for which a separate set of limitations and performance data exists in the AFM. The occurrence aircraft had a de-rated thrust of 46 300 pounds (JT9D-7 dry) and was referred to as "Rating II (RTG II)" in MK Airlines Limited documentation. Reduced take-off thrust is a thrust setting up to 25 per cent less than the maximum or de-rated take-off thrust. A reduced thrust setting is not restrictive in that it allows the flight crew to use maximum thrust at any time during the take off, if desired.
The MK Airlines Limited OM stated that, when setting take-off thrust, the operating crew must advance thrust levers to 1.10 engine pressure ratio (EPR),9 check that engine indications are stable and symmetrical, then advance thrust levers to approximately 1.20 EPR and call for "max thrust"10 to be set by the flight engineer.
According to the B747 AFM, Section 4, Performance, the stall speed for flap 20, at idle power and 353 800 kg, is 133 knots calibrated airspeed (KCAS). The stall speed is based on the aircraft in-flight and out-of-ground effect. The expected minimum unstick speed (Vmu)11 for the Halifax International Airport configuration was determined to be approximately 150 ± 2 KCAS. The B747 has an over-rotation stall warning system that activates a control column shaker during take-off when the rate or angle of rotation is excessive. The warning is deactivated when a body landing gear leaves the runway. Two stall warning systems are activated when the nose gear leaves the runway. Control column shaker was not a recorded value on the flight data recorder (FDR).
The B747 AFM indicated that, for the pressure altitude and airport temperature at the Halifax International Airport at the time of the occurrence, an EPR setting of 1.60 was required for maximum thrust, with a maximum reduction of 0.21 EPR for reduced thrust. The de-rated maximum thrust EPR setting was 1.43, with a maximum reduction of 0.14 EPR for reduced thrust.
Climb power for the occurrence flight, derived from the MK Airlines Limited quick reference handbook (QRH), was 1.33 EPR. During a reduced thrust take-off, some pilots at MK Airlines Limited would set climb EPR rather than take-off EPR if the climb EPR was the higher value. Go-around power from the QRH was 1.52 EPR.
According to the aircraft manufacturer, the B747-200 lower aft fuselage will contact the ground at a pitch attitude of 11.1º with static body gear oleo compression, and 13.1º with the body gear fully tilted and the oleos fully extended. The MK Airlines Limited OM indicated that the normal target pitch attitude for rotation is 12º with a rotation rate of 2º to 3º per second; lift-off should occur at approximately 10º pitch attitude. The manufacturer has determined that, for every five knots of airspeed below rotation speed (Vr),12 the angle of attack must be increased by 1º to gain the equivalent amount of lift during the rotation.
The 0600 Halifax International Airport weather was as follows: wind 250º true (T) at five knots, visibility 15 statute miles (sm), overcast clouds at 1700 feet above ground level (agl), temperature 10ºC, dew point 9ºC, and altimeter setting 29.67 inches of mercury (in Hg). The weather issued at 0700 was as follows: wind 260ºT at six knots, visibility 15 sm, overcast clouds at 1800 feet agl, temperature 10ºC, dew point 9ºC, and altimeter setting 29.67 in Hg. The airport's terminal area forecast corresponded to the actual weather.
At the time of the accident, the crew was using visual references for the take-off and was not relying on ground-based navigation aids. No discrepancies were discovered with the aids to navigation.
All communications between the Halifax International Airport air traffic controllers and MKA1602 were normal, and there were no deviations from published procedures. There were, however, some problems with the Aircraft Rescue and Fire Fighting (ARFF) communications (See Section 1.14.3 of this report).
The Halifax International Airport is located at latitude 44º52.85' N and longitude 063º30.52' W, at an elevation of 477 feet. It is a certified aerodrome operated by the Halifax International Airport Authority (HIAA) on land leased from Transport Canada (TC). Runway 24 was in use at the time of the accident. It is oriented 234º magnetic (M), constructed of asphalt and concrete, and is 8800 feet long by 200 feet wide. Runway 24 has a published take-off run available of 8800 feet and a clearway of 1000 feet, providing a take-off distance available of 9800 feet.
Just before impact, the aircraft severed a power cable and several telephone cables supplying the airport. Four diesel generators with auto-start capability, available to provide backup power to the airport power grid, started when the power cable was cut. Three of the generators supplied power to the airport grid; however, a circuit breaker tripped due to a power surge when the aircraft cut through power lines adjacent to the main crash site, preventing the fourth generator from supplying power. Approximately one hour after the accident, power from the fourth generator was restored when technical personnel manually reset the main circuit breaker. The control tower at Halifax International Airport was equipped with a separate stationary uninterruptible power unit and an independent backup power generator; consequently, there was no loss of electrical power to the tower.
The Halifax International Airport fire hall normally would receive backup power from two of the four generators mentioned above. The generator with the tripped circuit breaker should have powered a relay to permit operation of the following fire hall systems: bunkroom lights, vehicle bay lights, and the automatic opening of the vehicle bay doors. Because these systems were not powered, the firefighters had to respond in a darkened environment, and the vehicle bay doors had to be opened by pushing the manual door-open button at each bay. Because the door motors were powered by an operating generator, the doors then opened. The vehicle bay lights in the fire hall were "high-pressure sodium bulbs," which take approximately 10 minutes to reach full brightness; therefore, they would have been ineffective in a quick response scenario.
Had the fourth generator operated as expected, it would have taken 25 to 30 seconds for the bunkroom lights to come on, because of the time it would have taken for the fourth generator to reach full capacity. The fire hall had been equipped with self-contained battery-operated lights; however, when the emergency power generators were installed, these lights were removed.
In 2002, TC requested that NAV CANADA13 publish a slope of 0.17 per cent down for Runway 24 at Halifax International Airport in the Canada Flight Supplement and the Canada Air Pilot. TC's TP 312, Aerodrome Standards and Recommended Practices, Section 3.1.2.1, described how to calculate runway slope. Using TP 312, investigators from the Transportation Safety Board of Canada (TSB) calculated the slope for Runway 24 to be 0.19 per cent up. This error in direction and magnitude was not detected by NAV CANADA personnel before this information was published, nor was the error detected during subsequent reviews of these publications by the airport operator.
There are no standards for publishing slope values or slope changes for runways at Canadian airports, except that NAV CANADA documentation indicates that a slope of less than 0.3 per cent is not to be published.
Runway 24 has several slope changes. The two most significant are from the threshold of Runway 24 to the highpoint of the runway, which is 6975 feet from the threshold. The slope for this section is 0.24 per cent up. The slope for the remaining 1825 feet is 0.55 per cent down. The total absolute change in slope is 0.079 per cent up.
A review of non-Canadian aeronautical publications available to flight crews revealed conflicting information. One of these publications did not indicate any slope information for Runway 24. Another publication had the correct value and direction. A third described the slope for Runway 24 in two segments. The BLT runway information for Runway 24 was imported from a SITA14 data file on 19 September 2003 at 0952. It stated that Runway 24 had a slope of 0.08 per cent up and a field length of 8800 feet, plus 150 feet of paved overrun.
International Civil Aviation Organization (ICAO) Annex 15, Aeronautical Information Services, Part 3, specifies that the detailed description of runway physical characteristics for each runway is to include information on the slope of each runway and its associated stopways. Chapter 2 of ICAO Annex 4, Aeronautical Charts, Paragraph 2.17.1, states in part, "States shall insure that established procedures exist in order that aeronautical data at any moment is traceable to its origin so to allow any data anomalies or errors, detected during the production/maintenance phases or in the operational use, to be corrected." The Canadian Aviation Regulations specify that the operator of an airport shall review each issue of each aeronautical information publication on receipt thereof and, immediately after such review, notify the Minister of Transport of any inaccurate information contained therein that pertains to the airport.
An earthen berm, with a concrete slab on top to anchor the localizer antenna, was located 1150 feet from the end of Runway 24 on the extended centreline (see photos 2, 3 and 4). This berm was constructed in the fall of 2003 to support a new localizer antenna at a height necessary to meet ICAO localizer signal coverage requirements. The berm was 11.6 feet high, but since the terrain sloped downwards from the end of the runway, the concrete pad on top of the berm was in fact the same elevation as the end of the runway. The localizer antenna projected another 10 feet from the top of the berm. At the same time, a similar berm was constructed off the end of Runway 06 at a distance of 650 feet from the end of the runway. There are similar earthen berms in use at other airports in Canada, including one at Fredericton, New Brunswick, and several at Toronto/Lester B. Pearson International Airport, Ontario.
NAV CANADA submitted an Aeronautical Obstruction Clearance Form to TC on 27 August 2003 for the construction of both berms to support new localizer antennae. Approval was received on 08 September 2003. However, airport personnel raised a number of concerns when the berm on the approach to Runway 24 was first being constructed, primarily because it was thought to be a potential hazard. The HIAA corresponded with TC and requested clarification on whether the berm would affect the airport's certification. Based on an inspection of the berm by TC personnel, TC advised the HIAA in a letter dated 08 October 2003 that the berms for the new localizers on both Runway 06 and Runway 24 were not in conflict with airport certification standards.
In a follow-up letter from TC to the HIAA on 22 October 2003, TC stated, "Based on information supplied by NAV CANADA, we have determined that the subject localizers are in compliance with airport certification standards. Additionally, clearways are not affected and the existing TODAs [take-off distances available] will remain unchanged." The letter concludes, "Thus, from an airport certification perspective, we have no concerns about the installation of the new localizers on Runway 06 or Runway 24."
Airport certification standards are contained in TP 312. Each end of runways 06 and 24 had a clearway to ensure that there was an obstacle-free zone for departing aircraft. An obstacle-free zone comprises the airspace above the approach surface, inner transitional surfaces, and that portion of the strip bounded by these surfaces that is not penetrated by any fixed obstacle other than one that is required for air navigation purposes, is low mass and frangibly mounted. The HIAA did not list any obstacles, as defined in TP 312, for the departure paths for runways 06 and 24. The earthen berm was not considered an obstacle because it did not penetrate into the obstacle-free zone.
TP 312 uses the ICAO phraseology of "standard" or "recommended practice" to identify specifications considered to have a direct impact on the safety of flight from those that affect only operational efficiency. Only the standards contained in TP 312 are mandatory for the certification of Canadian airports; recommended practices are optional and might or might not be implemented. One of the recommended practices in TP 312 is to establish a runway end safety area (RESA). A RESA is defined as an area symmetrical about the extended runway centreline and adjacent to the end of the strip, primarily intended to reduce the risk of damage to an aeroplane undershooting or overrunning the runway.
According to TP 312 recommended practices, a RESA should extend from the end of a runway strip for as great a distance as practicable, but at least 90 m (295 feet). The runway strips for runways 06 and 24 at Halifax extend for 60 m (197 feet) beyond the threshold of each runway. The minimum distance specified for a RESA in the recommendations therefore would be at least 150 m (492 feet) at Halifax International Airport. The berms for the localizers for runways 06 and 24 are both located beyond these minimum recommended distances. There is no RESA published for the Halifax International Airport. ICAO considers a RESA to be a standard (ICAO Annex 14, Section 3.5.1) rather than a recommended practice.
The following automatic terminal information service (ATIS) broadcasts were issued during the time MKA1602 arrived and departed Halifax International Airport:
The cockpit voice recorder (CVR) was a Collins model 642C-1, part number 522-4057-010, serial number 1660, that was fitted in March 2004. The CVR was found under debris in its mounting bracket near its installed location (see Photo 5), and it had been exposed to fire and extreme heat for an extended period. The recording tape had melted; consequently, no CVR information was available to investigators. Although this model of recorder was not required to meet the more stringent fire test requirements that exist today, the conditions of extreme heat were such that the likelihood of any tape-based recorder surviving in those conditions is considered very low.
The flight data recorder (FDR) was a Sundstrand, part number 981-6009-011, serial number 2756, that was fitted in April 2004. It had a 25-hour recording capability and recorded a total of 107 parameters. The recording medium was Vicalloy tape. The FDR was found in the main cabin area forward of the wing root (see Photo 5). The FDR suffered impact and heat damage in the crash and the tape broke in two places. The FDR contained information from the previous six flights and good data for the accident flight. A small portion of data for the accident flight was not available because of the necessity to splice the tape where it had broken during the impact sequence.
The FDR data had several areas where data were lost due to signal distortion and dropouts. In some areas, the distortion was such that no recovery could be made. Data cycling causing dropouts was observed during the taxi segment, the initial portion of the take-off and the final 12 seconds of the recording. The data cycling was left as valid data to show this characteristic on the data plots (before and at the start of the take-off roll), even though the recorded data for the affected parameters were not valid. This cycling was tagged as invalid in the last 12-second segment of the flight to remove the dropouts from the data plots.
After push back, the aircraft began to taxi, the flaps were extended to 20º, and the horizontal stabilizer was set to 6.1 trim units,16 where it remained for the duration of the flight. The flight control checks were completed during the taxi. The aircraft entered Runway 24 at Taxiway Delta and backtracked to the threshold. The aircraft then made a 180º turn to the right and, upon lining up with the runway (234ºM), the thrust levers were advanced and a rolling take-off was comme
All the occupants were identified by DNA testing, and where dental records were available, they were used to verify the identity of the crew members. Forensic examination and toxicology tests did not indicate any physiological conditions or the presence of foreign substances that might have impaired the performance of any flight crew member.
The Halifax International Airport ARFF met the service standards specified in the Canadian Aviation Regulations, Part III, Section 323, Aerodrome and Airport Standards, which refer to aircraft fire fighting at airports and aerodromes. The ARFF at Halifax International Airport was last inspected by TC on 27 November 2003 and no deficiencies were found.
On 29 July 1997, a Mutual Aid Fire Fighting Agreement was signed between Halifax International Airport ARFF and the Halifax Regional Fire and Emergency Service. This type of agreement is often used by airports and municipal fire departments to outline the responsibilities of the parties involved where one agency requires the other's assistance. Although the agencies had the mutual aid agreement, there had only been limited opportunities for mutual inter-agency training.
On seeing the fireball of the aircraft, the Halifax International Airport tower controller activated the crash alarm. The airport ARFF units responded and arrived at the accident site approximately five minutes after the crash alarm sounded; the site was on airport property, but outside the airport perimeter security fence. The Halifax Regional Municipality (HRM) Fire Department, the Enfield Fire Department, Emergency Health Services and the Royal Canadian Mounted Police (RCMP) all responded within minutes of the accident. The RCMP established a security perimeter and controlled access to the site.
The HIAA Emergency Plan Manual contained a grid map of the Halifax International Airport that depicted the airport runways, taxiways, structures, roads, security fence, and the airport property boundaries. The map was divided into a numbered and lettered grid to permit rapid and clear identification of any response area. Copies of the grid map were posted in the airport fire hall, the ATC tower cab, the security operations centre and the emergency operations centre, and were carried in all airport emergency vehicles.
On hearing the crash alarm, the fire brigade captain in the airport fire hall contacted the air traffic controller over the crash phone to get confirmation that an emergency situation was in progress. The location of the accident site was described to him in general terms of the area and direction, which was normal controller practice; grid map coordinates were not used. When the airport firefighters departed the fire hall, they saw the fire and proceeded to the accident site.
The NAV CANADA air traffic controllers were provided training on how an aerodrome grid map is used to identify areas of an airport. In support of ARFF training at the Halifax International Airport, the air traffic controllers had used the grid map in the past to direct the response of the firefighters to different locations in the airport.
The 911 operator directed the other responding agencies to the accident site by relaying the general location description that was available. This resulted in some confusion as to the exact location of the accident site. There are no regulations, standards or local procedures that require the use of aerodrome grid maps for emergency response and none of the other responding agencies had copies of the Halifax International Airport aerodrome grid map available in their vehicles.
A TSB Safety Advisory was sent to TC following a 1998 accident (TSB report A98Q0192), on the subject of using grid maps for reliable and efficient direction during emergency responses to aircraft accidents. TC replied on 14 August 2000 to advise the TSB that the subject was to be addressed in a Notice of Proposed Amendment to the aerodrome safety regulations and standards. This change has not occurred to date.
Each ARFF vehicle was equipped with very high frequency, two-way radios, which were used to communicate with the air traffic controller on ground control frequency. The firefighters also had portable ultra high frequency radios that, because of a blind spot at the site, could not be used to communicate with the HIAA Emergency Operations Centre. These radios could have been used at the site for ARFF communication had they used the "simplex" mode, which allows direct communications at short distances.
The ARFF firefighters eventually had to communicate using hand signals until they were supplied with a portable trunk mobile radio from another response agency. The Halifax International Airport ARFF had a portable trunk mobile radio unit that could have been used to communicate with outside agencies such as the Halifax Regional Fire and Emergency Service dispatch facility and the RCMP. However, it had been left at the fire hall. The firefighters also attempted to use their individual cell phones to communicate with the Emergency Operations Centre, but the signal was unreliable.
The Halifax International Airport ARFF firefighters were the first on the scene and took command of the situation. Although they were responding outside the airport perimeter security fence, they suspected that the aircraft was still on airport property. The HIAA Emergency Plan Manual stated that the airport's ARFF was to be the lead agency in the event of an aircraft crash "on airport," and the municipal fire department was to be the lead agency in the event of an aircraft crash "off airport." Although the responsibilities for on or off airport crashes were described in the different documents, the actual boundary separating the two areas was not clearly defined. In several locations, the airport property limits extended outside the airport perimeter security fence.
When HRM Fire Department firefighters arrived at the accident site, they observed the aircraft to be outside the airport perimeter security fence; consequently, they assumed that they were the lead agency. A unified command post was established that included the HRM Fire Department, Emergency Health Services, the RCMP and, eventually, Halifax International Airport ARFF. This temporary confusion as to who was in command of the site did not cause significant problems. Some responders noted that there was a need for more inter-agency training.
Information on the number of persons and dangerous goods carried on board the aircraft was not readily available to air traffic controllers in the Halifax International Airport tower. Shortly after the accident, the air traffic controllers unsuccessfully attempted to contact the ground handling agency at the airport to learn the number of persons and dangerous goods on board MKA1602.
In accordance with ICAO's Technical Instructions for the Safe Transport of Dangerous Goods by Air, shipping documentation must accompany dangerous goods on board an aircraft. The MKA1602 flight crew had copies of the shipping documentation and copies were also available at Bradley International Airport, where the dangerous goods were loaded on the aircraft. Regulations only require that copies of shipping documentation be left at an airport where the dangerous goods have been loaded; therefore, authorities at intermediate airports are not aware of all the dangerous goods that might be on board an aircraft.
Approximately one hour after the accident, MK Airlines Limited operations staff contacted the Halifax International Airport control tower and informed the controller that there were seven crew members on board MKA1602. This information was immediately relayed to ARFF and to the HIAA Emergency Operations Centre. At 0840, MK Airlines Limited informed the ARFF that no dangerous goods had been loaded on MKA1602 at the Halifax International Airport. At 1700, about 10 hours after the accident, MK Airlines Limited sent a 30-page fax listing the dangerous goods that had been loaded on board the aircraft at Bradley International Airport. The goods included medical supplies, adhesives, paint, food flavouring, and motor vehicles.
When an aircraft in flight declares an emergency, controllers get the information regarding the number of persons and dangerous goods on board directly from the flight crew. In the event that it is not possible to communicate with the flight crew, the affected company should be able to provide this information, though it might take some time. However, rarely do controllers have the required contact information for the many airline companies.
The occupants were all located in the cockpit and upper deck rest area behind the cockpit. These areas were severely compromised during the impact and break-up of the aircraft. There was also an intense post-crash fire. The accident was not survivable.
The aircraft manufacturer was requested to provide analysis on the aircraft performance characteristics of MKA1602 during the take-off at Halifax International Airport. The manufacturer used two independent software tools to assess the take-off of MKA1602, using the FDR data from the accident and previous flights. The TSB and the United States National Transportation Safety Board (NTSB) reviewed the manufacturer's analysis. It was found that the performance characteristics of MKA1602 were consistent with that expected for normal operation. The simulation EPRs and recorded FDR EPRs were similar, further validating the simulation models for the take-off roll condition. Both software tools provided a consistent result that showed the actual weight of the aircraft to be reasonably near to that calculated by the TSB for the attempted take-off at Halifax International Airport.
The manufacturer's engineering analysis also provided a hypothetical flight path the aircraft would have taken if the berm had not been present. Limitations of the software and the many assumptions required to model such a take-off (where the aircraft lower aft fuselage is dragging) made it difficult to determine a clear result. However, assuming that the performance of the aircraft remained as it was before impact with the berm, it is considered likely that the aircraft would have stayed airborne, possibly contacting tree tops located 2000 feet beyond the end of the runway. The elevation of the top of the berm was the same as the end of the runway, and the tree tops were estimated to be at about the same elevation as the berm or slightly higher (see Photo 3).
The company began operations with a single DC-8 aircraft in 1990 as Cargo d'Or, using a Ghanaian air operating certificate (AOC). During this same period, the airline established an office in the United Kingdom near Gatwick Airport to facilitate general sales for the company. In 1993, the company invested in another airline in Ghana called Venus Air, and transferred the Cargo d'Or aircraft to the Venus Air AOC. Concurrently with the transfer, the name of the airline was changed to MK Airlines Limited. In November 1993, the commercial offices in Gatwick were moved to the present location at Landhurst, East Sussex. The enhanced communication and infrastructure potential at the new location facilitated improved in-house management functions, flight training, maintenance, crew scheduling, and operational control.
As the company expanded, more DC-8 aircraft were added to the fleet and more employees joined the company. Most of the new flight crew members were from southern Africa, and many had a military background and/or a previous connection with the managing director or other employees of MK Airlines Limited. The company philosophy was to provide people from that geographic area with employment opportunities that would not otherwise exist.
The first B747 aircraft was added to the company's fleet in 1999. At the time of the accident, the company was operating six DC-8 and six B747 aircraft. Over the last several years, the company had increased its fleet by approximately one aircraft each year to accommodate the growing demand for cargo capacity, which was increasing by approximately 30 per cent each year. At the time of the accident, the company employed about 450 people. However, it was reported by several flight crew members that there were crew shortages, especially on the B747. These shortages were due to company expansion, training demands, and crew retention issues (see Section 1.17.1.5 of this report).
At the time of the accident, MK Airlines Limited held a Ghana Civil Aviation Authority (GCAA) AOC (No. 16/18/2003), issued 22 December 2003 and valid until 31 December 2004. The AOC was granted for the purpose of public transport, passengers, mail, cargo, and aerial work with B747 and DC-8 aircraft. The company had an extensive intercontinental route structure with many of the routes being long, triangular patterns to best serve the market demands.
The MK Airlines Limited OM described how the company would manage its flight operations. In anticipation of Ghanaian parliamentary passage of the 2002 GCARs (see Section 1.17.2.2 of this report), the OM also described some programs that were not fully developed or implemented, such as the flight operations quality assurance and flight safety program.
In accordance with the MK Airlines Limited OM, the Operations Manager was responsible for ensuring that an adequate level of flight operations supervision was maintained. For up to two years before the accident, the Operations Manager position was filled by the B747 fleet captain in an acting capacity. In addition to his duties as the acting Operations Manager, his responsibilities as the B747 fleet captain required him to do a considerable amount of line flying. His line flying enabled him to exercise adequate supervision of operations and allowed crews to express their concerns and raise issues directly to him. However, some of the operations management responsibilities were not being fully carried out. For example, the OM was not being kept up to date, the supervision of flight and duty limitations was lacking in some areas, and consistent adherence to SOPs was not being assured.
Many of the MK Airlines Limited flight crew had similar backgrounds. Employees of the company had a familial approach to business, which permeated all levels, including the line crews and supervisory/management personnel. This familial environment resulted in both positive and negative consequences for the company. For example, on the positive side, it provided a strong sense of loyalty and commitment to the success of the company. On the negative side, it created an environment where managers and supervisors could have had difficulty ensuring that their "friends" adhered to company procedures and policies. For example, some supervisory pilots had noted occurrences of non-adherence to SOPs when they were non-operating crew members. These were not brought to the attention of the crew, nor were they reported to the company because of this familial relationship and their status on the flight. It was noted that several supervisory pilots had flown the MKA1601/MKA1602 flights, where the maximum allowable duty hours were exceeded with no action being taken.
Due to the nature of the non-scheduled air cargo operations and the routes that were being flown, there were ongoing, significant challenges faced by management and crews. These included departure delays, schedule changes, aircraft unserviceabilities, inhospitable destinations, and crew flight-time limitations. In this context, both management and crews occasionally felt it was appropriate to deviate from company policy and procedures to accomplish the mission. This was done believing that the risk in doing so was manageable.
The flight operations quality and flight safety program described in the OM was relatively new and had been somewhat slow in developing. The company wanted a program developed in house and one that reflected the company culture, rather than one that was "off the shelf." Company management reported that they had an open approach to flight safety and regarded it as being very important. Safety information was distributed quickly to crews through their computer-generated flight briefs. At the time of the accident, some components of the flight operations quality and flight safety program described in the MK Airlines Limited OM were not actually being carried out, or were only being partially carried out.
Although there was an occurrence investigation system and occurrence tracking software had been acquired, the database was still being developed. There was no confidential reporting system.
There was no flight operations quality assurance audit program in place for flight operations. However, the company had been assessing a number of different systems for the retrieval and analysis of FDR information.
MK Airlines Limited had been undergoing rapid expansion, especially in the B747 operation. The company had evolved from outsourcing almost all of its training to having a complete in house capability. At the time of the accident, the company was conducting approximately six B747 conversion courses each year. The company operated its own flight simulators, one B747 and one DC-8.
The company also had a comprehensive 40-hour flight crew indoctrination training program that was required for all newly hired flight crew.
The MK Airlines Limited OM, Part D - Training, Appendix A, contained a ground and simulator training course syllabus for the DC-8. There was no equivalent B747 ground and simulator course syllabus in the OM, Part D; however, there was a separate manual describing a ground training syllabus and a simulator training program.
The B747 classroom instruction was supplemented by practical application of the subject matter in the simulator. The simulator flight training syllabus was quite extensive; 14 four-hour sessions were provided with a proficiency check and instrument flight test conducted after these sessions.
A system of tracking the training required by different crew members and the filing of the different training records was in place; however, when TSB investigators requested training files of the occurrence crew, some documents were missing or were difficult to locate.
Training on new technology equipment and software, such as the BLT, was done by self-study and hands-on experience, using training material developed from the manufacturer's software manual. The information was distributed through notices to flight crews but had not been incorporated into the OM. There was no formal documentation to record an assessment of the individual's knowledge and competency using the equipment.
The company had a dedicated, experienced and knowledgeable group of supervisory pilots and flight engineers in its Training Department. One aspect stressed by the Training Department during training sessions and line checks was the strict adherence to SOPs at all times. Generally, most employees felt the training was adequate and the SOPs were appropriate for the operation. The operating philosophy among most flight crew was that SOPs were to be followed unless there were justifiable extenuating circumstances.
A significant number of MK Airlines Limited employees, particularly flight crew members, lived in southern Africa. Because of the company's business locations and route structure, employees were separated from their families for weeks at a time when on duty. With the political and social unrest in some of these areas, there was the potential for harm to come to their families when the employees were away. There were several examples cited where employees' families had experienced incidents of home invasion and/or personal attack. This was identified as a source of stress within the company.
In an effort to improve working conditions at MK Airlines Limited, the managing director had requested, some time before the accident, that the captain of MKA1602 submit a letter on behalf of the crews, listing some general concerns and suggestions of other flight crew. The letter was submitted shortly before the accident, and the company voluntarily supplied it to the TSB investigators. The letter indicated concern about recent increases in the number of pilots leaving the company and suggested that a new compensation package should be put in place to provide a more stable financial situation for flight crew members. The letter also indicated that there were not enough crews per aircraft. As well, it discussed the uncertainty of life for those living in southern Africa, indicating that the lengthy periods away from home increased stress and contributed to flight crew members looking at other employment options. The letter mentioned that inexperienced operational support personnel, combined with pressure from the Commercial Department, were causing crew scheduling difficulties.
Other company employees reported that there was a consistent shortage of B747 flight crew and they were required to spend lengthy periods away from home. To address a crew shortage in the past, the company had hired flight crew members from Argentina on contract to supplement its DC-8 operation.
A review of the technical records indicated that all requirements of the approved maintenance program had been completed on the accident aircraft in accordance with the variation/tolerance approved by the GCAA.
Since 1991, the GCAA has invested in personnel, training and equipment to help ensure conformity with ICAO standards and recommended practices (SARPs). In 1993, the GCAA requested assistance from Canada in the form of a safety review by TC. Following this review, a report was provided to the GCAA that identified areas for improvement, including a rewrite of the regulations and development of oversight guidance for inspectors. Several Canadians were contracted in 1995 to assist the GCAA in this effort. A new set of regulations came into effect in 1995.
The regulations that were in force at the time of the accident were the 1995 GCARs. In 1997, ICAO identified that these regulations needed to be updated.
ICAO conducted an audit of the GCAA in April 2001 and noted in its report that the GCAA's corrective action to the audit findings was generally satisfactory. The audit also determined that the newly drafted GCARs (referred to as the 2002 version) were in compliance with most of the SARPs, but had not yet been approved by the Ghanaian parliament.
An audit follow-up was conducted in May 2003 to determine the progress made on the corrective action. It was noted that some progress had been made, but the GCAA's regulatory efforts were being hampered by the ongoing delay in bringing the new regulations (2002 GCARs) into force. ICAO noted at that time that Ghana had five AOCs issued and there were two approved maintenance organizations. On 11 November 2004, the parliament of the Republic of Ghana approved the Civil Aviation Act, 2004, which brought into law the 2002 version of the regulations.
The GCAA was asked to provide all records of all inspections, audits, and correspondence related to MK Airlines Limited for the two-year period before the accident at Halifax International Airport. The GCAA operations inspection file for MK Airlines Limited was reviewed to determine the actual frequency of inspections and to assess the handling of any safety deficiencies identified by the GCAA. It was concluded that the actual inspections conducted during the two years before the accident were below the minimum frequency of about 20 inspections indicated in the inspector's handbook. The inspection frequency of MK Airlines Limited had been decreased due to increased vigilance and the inspection of another registered Ghanaian operator.
A GCAA base inspection in September 2003 identified areas in the MK Airlines Limited OM that needed revision; however, many other areas were not identified as being out of date, not being followed, or in conflict with the regulations. An example of OM conflict with regulations was the practice of one pilot leaving the cockpit for prolonged periods during a flight. Although this was identified as a deficiency during a GCAA in-flight inspection, and formally recorded, the GCAA inspector was apparently unaware that the MK Airlines Limited OM, Section 8.3.10.1 (revised in 2001), provided for a flight crew member leaving his assigned station for an agreed purpose and period with the permission of the captain. It must be noted that OM, Section 8.3.10.1, contradicted OM, Section 7.4, which prohibited the pilot from leaving his duty station for a prolonged period.
Although the GCAA was asking for compliance with the 2002 GCARs, the company felt that the 1995 version was still in effect, and this might have led to different interpretations. The 1995 version of the GCARs stated that, if the aircraft is required to carry two pilots, the commander shall cause both the pilots to remain at the controls for only the take-off and landing. The GCAA reported that it was not aware of the rest, duty and flight time scheme in use by the company at the time of the accident, even though it had been in use and included in the company OM for two years.
The 1995 GCARs indicated that periodic checks were to be carried out by the GCAA, and that these checks were to be done in conjunction with supervisory visits. These regulations did not specify the frequency or quantity of these checks or visits. The GCAA airworthiness inspectors had participated in the base inspection of MK Airlines Limited Landhurst facilities in September 2003. Additionally, it was reported that the GCAA carried out on-site visits for the annual renewal of each aircraft's Certificate of Airworthiness. During these visits, the GCAA would carry out an inspection that focused on examining the aircraft and reviewing the applicable technical documents. With a fleet of 12 aircraft, the GCAA would be at MK Airlines Limited facilities at least 12 times per year to carry out inspections.
Although some discrepancies were noted during the airworthiness review of MK Airlines Limited, overall, it appeared that the GCAA was providing an adequate level of airworthiness oversight. In a letter from the GCAA to MK Airlines Limited after an airworthiness audit, it was noted that amendments that had been incorporated in the maintenance control manual and the minimum equipment list had not been referred to the authority for prior approval. The GCAA did not receive a corrective action plan promised by MK Airlines Limited from the base inspection in September 2003.
TC's Foreign Inspection Division conducted a base inspection of MK Airlines Limited operations in the United Kingdom between 15 August and 26 August 2002. The closing paragraph of the base inspection report stated that the company would be issued a Canadian Foreign Air Operators Certificate upon receipt of an acceptable corrective action plan that addressed the findings of the inspection. MK Airlines Limited submitted a corrective action plan in October 2002. On 20 December 2002, TC's Foreign Inspection Division granted MK Airlines Limited Canadian Foreign Air Operators Certificate F 10326. The Division had some concerns about issuing the certificate because of MK Airlines Limited accident history (see Section 1.18.6 of this report). However, the Division was impressed by MK Airlines Limited management, the timeliness and content of the corrective action plan, and the quality of feedback from the United Kingdom Civil Aviation Authority (CAA). Contributing to the confidence of the decision was the FAA's assessment that the GCAA was a Category 1 regulatory authority.
The FAA established the International Aviation Safety Assessments (IASA) program in August 1992 to assess the ability of foreign civil aviation authorities to ensure compliance with ICAO's SARPs. There are two IASA safety ratings with regard to the SARPs: does comply (Category 1) and does not comply (Category 2). Specifically, the FAA determines whether a foreign civil aviation authority has an adequate infrastructure for international aviation safety oversight as defined by ICAO standards.
The basic elements that the FAA considers necessary include the following:
In 1996, the FAA assessed the GCAA as having a Category 1 safety rating.
On 02 June 2003, the FAA granted MK Airlines Limited authority to operate in the United States by issuing Operations Specification ZM0F869F. As part of the FAA's oversight, periodic ramp inspections were conducted on MK Airlines Limited aircraft. In July 2004, MK Airlines Limited was placed on a special emphasis list. This list is issued semi-annually to identify foreign air carriers that are to be watched. The list also includes countries with a Category 1 Civil Aviation Authority, where the FAA has concerns. In September 2004, a ramp inspection of an MK Airlines Limited aircraft resulted in a decision to increase surveillance of the company's operation. A ramp inspection of an MK Airlines Limited DC-8 in the United States following the accident in Halifax identified several deficiencies, and on 29 October 2004, the FAA informed the company that its Operations Specification was cancelled; no specific reason was stated.
In December 2004, the FAA conducted a reassessment of the GCAA and, on 30 April 2005, it announced publicly that Ghana had failed to comply with ICAO standards. As a result, Ghana's safety rating was lowered to Category 2.
Within the United Kingdom, the Department for Transport will issue a permit to a foreign-registered airline to operate into the United Kingdom if the airline has all the relevant approvals from its regulating authority. Where the Department has reason to believe that an airline or aircraft might not comply with international standards, it can arrange for that airline's aircraft to be inspected by the CAA. Where the CAA finds a matter requiring attention, it will be raised with the aircraft crew, airline, and/or foreign authority as appropriate. The Department for Transport relies on the country of registry to carry out effective and ongoing oversight of the company. The Department had issued permits to MK Airlines Limited, and the United Kingdom CAA had conducted a number of ramp inspections on MK Airlines Limited before the accident, without any significant findings.
At the time of the accident, MK Airlines Limited was using the Boeing Laptop Tool (BLT) for determining performance calculations. The BLT is a Microsoft Windows®-based software application used to calculate take-off performance data, landing performance data, and weight and balance information (see Figure 2). The performance data in the software are a digitized form of the approved B747 AFM.
The weight and balance data were supplied by and built into the software by MK Airlines Limited, and Boeing provided training to the MK Airlines Limited software administrator. The MK Airlines Limited BLT software administrator was responsible for setting up the weight and balance page for each specific aircraft and for supplying the airport database for the BLT. Boeing did not approve or review the work that the company administrator had done to the BLT weight and balance page. The company administrator had the option to lock out the weight and balance page in the BLT to prevent crews from using it; however, MK Airlines Limited decided not to lock out the weight and balance page in order to leave the page as a cross-check against the loadmaster's manual calculations. MK Airlines Limited had also received a current Boeing Administrator's Guide, with comprehensive instructions and the latest revisions to the software.
The software version in effect at the time of the accident was version 2.69r, effective 24 May 2004. Each B747 in MK Airlines Limited's fleet had been equipped with a laptop computer with a touch screen and printer stored in the upper deck. The BLT software installed on each laptop was aircraft-specific because of the differences in engines and weight and balance among aircraft.
The weight and balance feature of the BLT software allowed a user to perform basic weight and balance functions, including the calculation of the stabilizer trim setting for take-off. This software feature was an option that MK Airlines Limited had activated.
The weight and balance feature was accessed via a dedicated button on the main input dialogue screen, which would bring up a weight and balance summary page. The user could enter passenger weights, cargo zone weights, and fuel for the flight. When that information was entered, the take-off and landing weight, based on those inputs, was updated at the bottom of the summary page. Once updated, the estimated take-off weight would be passed back to the planned weight field on the main input dialogue screen and would automatically overwrite any entry in the planned weight field, without any notification to the user.
When the MK Airlines Limited beta version19 of the BLT was ready, the Information Technology Department and the B747 Training Department began in-house testing, comparing the AFM performance section as a cross-check of the BLT using different samples of airports, altitudes, and temperatures. Differences were noted and corrected in conjunction with Boeing. The BLT was then given to the B747 Training Department instructors to begin training crews in its use. Information on the BLT was distributed to flight crews in the form of newsletters and notices to flight crews.
On 09 February 2004, the MK Airlines Limited B747 chief training pilot issued a Notice to Flight Crew to the B747 flight crew (including loadmasters) on the subject of the BLT. It stated the following:
Please find attached the Performance section and relevant QRH pages. Please take the time to study these for when the BLT program is put onto the onboard computers. The BLT will eventually replace the Airport Analysis Charts.20
This Notice to Flight Crew had a detailed 46-page manual attached on the use of the BLT to calculate performance data, which was issued as an amendment to the company OM, Part B, for the B747-200 series, Section 4, Performance. The self-study training material was to be read by the pilots, and they were encouraged to practise using the laptop on board each aircraft. If the pilots had questions or comments about the BLT, they were to be sent to the company. Line training captains conducting line checks provided some training, and some crews received BLT training during their regular recurrent aircraft training at Landhurst.
Most of the MK Airlines Limited flight crew members did not receive any formal training on the BLT, and there was no method to evaluate and record if individuals had become competent using the BLT by the end of the self-study training period. Company Training Department and management personnel were aware that some pilots were not comfortable using personal computers. No additional general computer training was offered to the flight crews. However, according to the management of MK Airlines Limited, none was ever requested.
The 46-page BLT manual issued 09 February 2004 had two references to a BLT feature regarding the planned weight dialogue box on the performance page. Whatever weight appeared in the dialogue box would be overwritten automatically with the estimated take-off weight from the weight and balance page, when the user moved from the weight and balance page to the performance page. On 29 March 2004, MK Airlines Limited issued a second Notice to Flight Crew for the B747 fleet, informing the crews that the BLT software had been installed on all aircraft computers and was approved for calculating performance data.
This two-page Notice to Flight Crew asked that the crews use the procedure as written to complete take-off data cards. The first page was an instruction to pilots, while the second page was an instruction to loadmasters. It stated in the loadmaster instructions that, "when closing the weight and balance page, the take-off weight as listed in the weight and balance page will now appear in the planned take-off weight block." There was no mention of this feature in the instructions to pilots.
This Notice to Flight Crew also requested that the flight crew members take time to read the BLT manual. It could not be determined if the occurrence flight crew read the BLT manual issued in February, or the simplified instructions issued in March. Reports from other MK Airlines Limited flight crews indicated that the operating captain was not comfortable using the BLT, while the first officer had been observed using it.
On 12 August 2004, MK Airlines Limited issued a Notice to Flight Crew, which stated the following:
Airport (runway) analysis charts will be removed from all aircraft libraries in the near future. All performance calculations are to be performed using the BLT. In the unlikely event that no BLT computers are working then please revert to Volume 2 for the calculations and make sure a MAX THRUST Take Off Data Card is used.
When the BLT software was opened, the introduction page presented the user with the option of two engine ratings for calculating take-off performance data: maximum take-off power using JT9D-7Q engine performance, identified by the aircraft registration (9G-MKJ), or de-rated power/rating II (RTG II) using the JT9D-7 dry engine certified performance, identified by RTG II.
Once the screen of the appropriate power rating is selected, the user inputs the airport and atmospheric data. The user then selects the "calculate" button and the BLT will indicate the maximum take-off weight for that runway and the EPR setting for maximum thrust for that power rating. The maximum thrust take-off performance data are displayed on the upper right of the screen, and the reduced thrust take-off performance data are displayed in the lower right side of the screen. The performance data on the right of the screen also include the aircraft weight on which the data were based. The user then transfers the appropriate data to a take-off data card (see Figure 3).
Based on the atmospheric conditions at Halifax International Airport at the time of the accident and for a take-off on Runway 24, the BLT would have indicated that the maximum weight for a maximum thrust (JT9D-7Q engine) take-off would be 355 230 kg, and the maximum weight for a RTG II take-off would be 321 580 kg. The maximum weight for a reduced thrust take-off using the JT9D-7Q engine would be 346 513 kg, and the maximum weight for a reduced thrust take off using RTG II would be 315 058 kg.
The BLT take-off performance page would have indicated that, for a take-off weight of 350 698 kg (the weight taken from the mass and balance sheet left behind by the flight crew), using maximum thrust, the EPR setting would have been 1.60 and the take-off speeds, respectively, would have been: V1 (take-off decision speed)=149 knots, Vr (rotation speed)=161 knots, and V2 (take-off safety speed)=171 knots. At that weight, the BLT would not have provided reduced thrust performance data, including an EPR setting.
The BLT take-off performance page would have indicated that, for a take-off weight of 353 800 kg (the estimated actual weight), using maximum thrust, the EPR setting was 1.60 and the take-off speeds, respectively, would have been: V1=150 knots, Vr=162 knots, and V2=172 knots. At that weight, the BLT would not have provided reduced thrust performance data, including an EPR setting. The AFM provided comparable values.
If the RTG II option had been selected, any weight input into the planned weight box greater than 321 580 kg would have given the following warning: "Planned weight exceeds max allowable take-off weight of 321 580 kg."
An EPR setting of 1.33 could have been obtained using the BLT and RTG II, reduced thrust with a minimum weight of 285 000 kg. At that weight, the take-off speeds, respectively, would have been: V1=137 knots, Vr=145 knots, and V2=151 knots.
Using the atmospheric conditions at the time of take-off on Runway 06 at Bradley International Airport, the BLT take-off performance page would have indicated that, for a take-off weight of 239 783 kg, using RTG II, reduced thrust, the EPR setting would have been 1.30 and the take-off speeds, respectively, would have been: V1=128 knots, Vr=128 knots, and V2=137 knots.
Using the atmospheric conditions at the time of take-off on Runway 24 at Halifax International Airport and a take-off weight of 239 783 kg, the BLT take-off performance page (see Figure 6) would have indicated the following:
Engine | Thrust | EPR | V1 | Vr | V2 |
---|---|---|---|---|---|
JT9D-7Q | Maximum | 1.60 | 130 | 130 | 145 |
JT9D-7Q | Reduced | 1.40 | 130 | 130 | 137 |
JT9D-7 Dry | Maximum | 1.43 | 120 | 124 | 138 |
JT9D-7 Dry | Reduced | 1.30 | 123 | 129 | 137 |
The aircraft's landing weight at Halifax International Airport was approximately 227 000 kg. For the atmospheric conditions at the time of landing on Runway 24, the BLT indicated that the aircraft would require a landing reference speed (Vref)21 of 133 knots. In accordance with MK Airlines Limited normal flap extension schedule, the airspee